فهرست مطالب

مهندسی بیوسیستم ایران - سال پنجاه و یکم شماره 1 (بهار 1399)

نشریه مهندسی بیوسیستم ایران
سال پنجاه و یکم شماره 1 (بهار 1399)

  • تاریخ انتشار: 1399/01/28
  • تعداد عناوین: 15
|
  • عبدالله ادیب زاده، حسن ذکی دیزجی*، ناهید عقیلی ناطق صفحات 1-10

    نیشکر یکی از مهمترین گیاهان صنعتی است که در ایران اولین منبع تولید شکر (در حدود 40 الی50 درصد) است. صنعت شکر با توجه به تامین انرژی روزانه شهروندان از نقشی کلیدی در میان صنایع مختلف کشور برخوردار می باشد. شکر علاوه بر مصارف خانگی دارای اهمیت ویژه ای در صنعت غذا به دلیل خاصیت شیرین کنندگی و حجم دهندگی می باشد. مقدار قند نیشکر اغلب در حد 15-10 درصد و در مواردی تا 17 درصد می باشد. جهت شروع برداشت عوامل مختلفی نظیر واریته و تاریخ کاشت و یا برداشت سال گذشته مهم می باشد. از طرفی نیشکر قابل انبار کردن و ذخیره سازی در کارخانه نیست و مواد قندی آن سریعا در اثر نگهداری تجزیه می گردند و وزن نیشکر در اثر از دست دادن رطوبت کم می شود. هرچه نیشکر تازه تر به مصرف کارخانه برسد، بهتر خواهد بود. لذا از دستگاه بینی الکترونیک استفاده گردید تا با آزمایش شربت های نیشکر و ارتباط آن با بوهای متصاعد شده از آن بتوان نوع واریته را جهت برداشت به موقع نیشکر تشخیص داد. از مزارع نمونه نیشکر چهار رقم مطرح (CP57، CP69، IRC99-02، و CP48) انتخاب گردیدند. از روش های تحلیل تفکیک خطی (LDA)، تجزیه مولفه های اصلی (PCA) و شبکه های عصبی (ANN) برای تشخیص واریته های مختلف نیشکر استفاده شد. نتایج نشان داد که هر سه روش دقت بالایی در کلاس بندی رقم دارند. اما روش های LDA و PCA نسبت به روش ANN نتایج بهتری داشتند. به طوری که برای تشخیص واریته های نیشکر دقت طبقه بندی آن ها به ترتیب 98.33% ، 97% و 96.7% بود. نتایج به دست آمده توانایی بالای ماشین بویایی را در تمایز بین واریته های نیشکر نشان داد که می توان این سامانه را به عنوان یک ابزار سریع و کم هزینه در صنعت نیشکر به کار برد.

    کلیدواژگان: نیشکر، ماشین بویایی، تحلیل تفکیک خطی، شبکه عصبی و تجزیه مولفه های اصلی
  • ناهید حسنکی، یعقوب منصوری*، عباس عساکره صفحات 11-21
    هدف از این پژوهش، مطالعه وضعیت مصرف انرژی در کارخانه تولید شکر کشت و صنعت کارون و بررسی فنی و اقتصادی پتانسیل استفاده از باگاس به جای گاز طبیعی است. اطلاعات مورد نیاز بر اساس سال زراعی 1395-1396 جمع آوری گردید. نتایج نشان داد مصرف سالیانه انرژی برق و حرارت در کارخانه ی شکر کارون به ترتیب 24 گیگاوات ساعت و 83/890 تن بخار (حرارت) بوده است که موجب انتشار 195 تن آلاینده هوا با هزینه اجتماعی 79/24 میلیارد ریال شده است. نرخ گرمایی خالص نیروگاه شرکت 15/27 مگاژول بر کیلووات ساعت محاسبه شد. با استفاده از 270 هزار تن باگاس مازاد، پتانسیل تولید 14/15 درصد کل انرژی کارخانه شکر و کاهش 29 هزار تن آلاینده در سال وجود دارد. بر اساس سال پایه 1395، نرخ بازده سرمایه باگاس سوز کردن دیگ بخار شرکت، 95/88 درصد و دوره بازگشت سرمایه کمتر از دو سال با صرفه جویی 60/30 میلیارد ریال در سال به دست آمد.
    کلیدواژگان: انرژی، باگاس، زیست محیطی، نرخ بازگشت سرمایه، نیشکر
  • سید جواد سجادی، سلیمان حسین پور*، شاهین رفیعی صفحات 23-35
    افلاتوکسین B1 (AFB1) سمی ترین گروه آفلاتوکسین هاست که باعث آلودگی محصولات کشاورزی شده و اثرات مرگ باری بر سلامت انسان دارد. تشخیص AFB1 در مواد غذایی و خوراکی توسط بیوسنسورها سریع، کم هزینه و دقیق است. در این مقاله به مدلسازی و شبیه سازی واکنش های شیمیایی در بیوسنسور پتانسیومتری AFB1 جهت تعیین ثابت های  بهینه نرخ واکنش پرداخته شده است. شبیه سازی واکنش های شیمیایی توسط نرم افزار COMSOL و بهینه سازی ثابت هاینرخ واکنش توسط شبکه عصبی مصنوعی و الگوریتم ژنتیک انجام شد. علاوه بر آن شبکه عصبی مصنوعی به عنوان تابع هدف مورد استفاده در الگوریتم ژنتیک به کار رفت. داده های تولید شده در مرحله شبیه سازی جهت آموزش و ارزیابی عملکرد شبکه عصبی مورد استفاده قرار گرفتند. نتایج بدست آمده نشان داد مدل COMSOL در مقایسه با داده های تجربی، پاسخ بیوسنسور را با MAPE برابر با 1023/0 % شبیه سازی کرد. شبکه عصبی مصنوعی آموزش داده شده با ساختار 1-48-5 نیز قادر به پیش بینی پاسخ بیوسنسور با MAPE برابر با 7074/0 % ، 9458/0 % ، 7473/0 % و 7492/0 % به ترتیب برای داده های گروه آموزش، اعتبار سنجی، آزمون و کل داده ها بود. نتایج بهینه سازی ثابت های  نرخ واکنش توسط الگوریتم ژنتیک و شبکه عصبی مصنوعی نشان داد شبیه سازی پاسخ بیوسنسور AFB1 با استفاده از شبکه عصبی مصنوعی و پارامتر های ورودی انتخاب شده توسط الگوریتم ژنتیک دارای کمترین خطای MAPE برابر با  0026/0 % در پیش بینی میزان مهار آنزیم AChE توسط AFB1 است.
    کلیدواژگان: بیوسنسور، مدلسازی، کامسول، شبکه عصبی مصنوعی، الگوریتم ژنتیک
  • رضا ترکاشوند، رضا امیری چایجان*، علی قاسمی صفحات 37-49

    در این پژوهش اثر متغیرهای موثر محتوای رطوبتی، اندازه ذرات، نوع ماده چسبان و شکل قرص بر خواص کیفی (اختلاف شاخص های رنگی ∆L*،∆a* و ∆b*) ، فیزیکی (چگالی واحد و چروکیدگی) و مکانیکی (نیروی نفوذ) قرص فشرده تولید شده از پودر گوجه فرنگی مورد مطالعه قرار گرفتند. متغیرهای مستقل شامل سه سطح اندازه ذرات پودر گوجه فرنگی (ذرات کوچک تر از mm 3/0، ذرات بین 3/0 تا mm 75/0 و ذرات بزرگتر از mm 75/0)، سه سطح محتوای رطوبتی (18، 36 و d.b.54%)، سه شکل قرص فشرده (کروی، استوانه ای و مکعبی) و سه نوع ماده چسبان (شربت فروکتوز 55 درصد، آب و شکر) بودند. برای بهینه سازی از روش سطح پاسخ و طرح دی- اپتیمال استفاده شد. نتایج نشان داد که نقطه بهینه برای قرص های فشرده شده متشکل از پودر گوجه فرنگی با محتوای رطوبتی % d.b. 4/33، اندازه ذرات mm 3/0، نوع ماده چسبان فروکتوز و شکل قرص استوانه ای که در آون در دمای °C60 خشک شده بودند به دست آمد. تحت این شرایط، شاخص مطلوبیت 826/0 محاسبه شد و مقدار بهینه متغیر های مستقل∆L*، ∆a*و ∆b*، نیروی نفوذ، چگالی واحد و چروکیدگی به ترتیب برابر با 89/35، 6/51، 23/23، N 2/267 ، kg/m3 2299 و 12/2 درصد به دست آمدند.

    کلیدواژگان: بهینه سازی، قرص فشرده گوجه فرنگی، محتوای رطوبتی، اندازه ذرات
  • محمد آهمند، فاروق شریفیان*، علی محمد نیکبخت، وحید رستم پور، ادریس رحمتی صفحات 51-61
    در تحقیق حاضر تحلیل انرژی و اکسرژی شبیه ساز کلکتور خورشیدی صفحه تخت مجهز به ناهمواری Inclined Broken Rib بر اساس داده های تجربی در یک مدار باز و همچنین بهینه سازی شرایط کارکرد سامانه صورت گرفته است. آزمایش ها در نه سطح دبی جرمی (kg/s 03/0، 04/0، 05/0، 06/0، 07/0، 08/0، 09/0، 10/0 و 11/0)، پنج سطح شار حرارتی (W/m^21000، 1100، 1200، 1300 و 1400) و سه سطح دمای هوای محیط (°C 20، 25 و 30) انجام شد. نتایج نشان داد که بیشترین و کمترین مقدار بازده انرژی به ترتیب در تیمارهای با دمای محیط °C20 و 25 ، دبی جرمی  kg/s11/0 و شار حرارتی  W/m^21000 و 1400 با مقادیر 8/49 و 3/0 درصد به دست آمد. همچنین بیشترین و کمترین مقدار بازده اکسرژی به ترتیب برابر 75/5 و 607/0 درصد در تیمارهای با دمای محیط °C20 و 30 ، دبی جرمی  kg/s11/0 و 03/0 و شار حرارتی  W/m^21000 محاسبه شد. روش سطح پاسخ برای بهینه کردن شرایط کارکرد جمع کننده خورشیدی به کار گرفته شد. شرایط بهینه در دمای محیط °C20، نرخ جریان جریان جرمی  kg/s11/0 و شار حرارتی  W/m^21000 به دست آمد. در این شرایط بهینه، بازده انرژی و اکسرژی به ترتیب 08/42 و 76/5 درصد با مطلوبیت 92/0 به دست آمد.
    کلیدواژگان: بازده ی اکسرژی، بازده انرژی، انرژی خورشیدی، روش سطح پاسخ
  • احمدرضا صالحیون، محمد شریفی*، مرتضی آغباشلو، حمید زیلویی، سعید مفتح صفحات 63-76

    هضم بی هوازی به منظور تولید بیوگاز روشی اثبات شده برای تولید انرژی تجدیدپذیر از زباله های جامد شهری است. در این تحقیق اجزای بخش آلی زباله جامد شهری در کلان شهر کرج تعیین شد. اجزاء بخش آلی زباله در پنج دسته ضایعات میوه، چربی و پروتئین، نشاسته، سبزیجات، و سلولزی در دو فصل زمستان و تابستان رصد شد. سپس یک نمونه که بیانگر متوسط مقدار اجزاء زباله باشد سنتز شده و عملکرد بیومتان، شاخص های هضم پذیری و پارامترهای مدلسازی سینتیک تولید بیوگاز در آزمون وعده ای در دمای مزوفیل و در دو سطح غلظت 8 و 15% TS بررسی گردید. بیشترین جزء در بخش آلی زباله به ضایعات میوه و سبزی با مجموع 9/62 و 6/70% در دو فصل زمستان و تابستان تعلق داشت. عملکرد بیومتان در 8 و 15% TS به ترتیب برابر با 2/385 و 2/289 L/kg VS و درصد متان 8/66 و 8/58 تفاوت معناداری داشت، اما تخریب VS با مقادیر 49/87 و 72/84% تفاوت معناداری نداشت. در نتیجه برای زباله تفکیک شده در مبدا هضم بی هوازی در TSهای پایین تر، نتایج بهتری نسبت به هضم خشک دارد. هضم پیوسته برای زمان ماند 30 روز برای داشتن تولید ویژه بیومتان و تولید حجمی بالا در شرایط پایدار قابل اجرا است.

    کلیدواژگان: هضم بی هوازی، زباله های جامد شهری، بیوگاز، آزمون وعده ای، مدلسازی
  • محمد شریفی*، لیلا بهروزنیا، سیدهاشم موسوی اول صفحات 77-87

    پسماند جامد شهری، به عنوان مواد دور ریز شناخته می شود و به طور روزافزون به این مواد در جهان اضافه می شود و سبب ایجاد مشکلات بیشتری از جمله آلودگی هوا و انتشار گازهای گلخانه ای در محیط زیست می شود. در نتیجه، نیاز به مدیریت درست و پایدار زباله توسط مدیران بیشتر احساس می شود. بر این اساس، تولید کمپوست، یکی از روش هایی است که علاوه بر کاهش آلایندگی، در کشاورزی نیز مورد استفاده قرار می گیرد. در شهر رشت نیز از روش تولید کمپوست برای مدیریت 400 تن زباله در روز بهره گرفته می شود. در فرایند تولید کمپوست، آلاینده هایی ایجاد می شود که باعث اثرگذاری در محیط زیست می گردد. در این مطالعه، برای ارزیابی چرخه زندگی، از نرم افزار سیماپرو، از روش CML-IA baseline V3.04 / World 2000 استفاده شد و 11 بخش اثر بررسی گردید و در نهایت نتایج نرمال سازی و وزن دهی شدند. واحد کارکردی در این مطالعه 400 تن کمپوست تولیدی به ازای یک روز در نظر گرفته شد. نتایج نشان داد که مقدار پتانسیل گرمایش جهانی برابر 103×28/4 kgCO2 eq. می باشد و بیشترین سهم را در این بخش اثر، آلاینده های مستقیم و حمل و نقل دارند. همچنین نتایج پس از نرمال سازی نشان داد که تولید کمپوست از زباله بیشترین اثر را به ترتیب در سمیت آب های آزاد و سمیت انسان دارد.

    کلیدواژگان: پتانسیل گرمایش جهانی، گازهای گلخانه ای، آلایندگی زیست محیطی، کمپوست
  • کامران ملکی مجد، داریوش زارع*، همایون امداد، سید مهدی نصیری، غلامرضا کریمی، خسرو جعفر پور صفحات 89-97
    در این تحقیق افت فشار حاصل از هوادهی مخزن حاوی ذرت به عنوان ماده متخلخل دانه ای در دو آرایش H و F کانال های ورودی هوا برای یک سیلوی آزمایشگاهی با استفاده از روش تجربی مورد مطالعه قرار گرفت. برای آرایش F افت فشار بیشتری نسبت به آرایش H مشاهده گردید. نتایج تجربی با معادله ارگان برازش داده شد و ضرایب ثابت اصلاح کننده معادله ارگان برای هر دو آرایش H و F تعیین شد. میزان تطابق معادله ارگان با داده های تجربی با استفاده از شاخص های آماری شامل ضریب تبیین، و درصد میانگین نسبی خطاها و مجذور میانگین مربعات خطاها مورد بررسی قرار گرفت که معادله ارگان با توجه به ضرایب اصلاح شده همخوانی خوبی با داده های تجربی داشت. با استفاده از معادله اصلاح شده ارگان شبیه سازی عددی جریان در داخل بستر ذرت با روش المان محدود در نرم افزار کامسول ویرایش 3/5 انجام گرفت، همچنین توزیع سرعت و خطوط جریان در داخل بستر مخزن نیز ارایه شد.
    کلیدواژگان: هوادهی مخزن، شبیه سازی عددی، مواد متخلخل، آرایش کانال ورودی
  • نسیم منجزی* صفحات 99-112

    هدف از این پژوهش، عارضه یابی، شناسایی و اولویت بندی عوامل موثر بر بهره وری آب در تولید نیشکر در استان خوزستان می باشد. در این تحقیق، برای محاسبه بهره وری آب از شاخص هایی چون CPD،BPD و NBPD استفاده شده است. با توجه به این که هدف تحقیق، اولویت بندی عوامل موثر بر بهره وری نهاده آب طی فرآیند تولید نیشکر است، عوامل شناسایی شده با استفاده از تکنیک تحلیل سلسله مراتبی (AHP) ارزیابی شدند. سپس در مدل سازی بهره وری نهاده آب و بررسی عوامل موثر بر آن از الگوریتم های تصمیم CART و CHAID استفاده شد. بر اساس نتایج تحقیق، متوسط شاخص های CPD، BPD و NBPD برای محصول نیشکر به ترتیب برابر 37/2 کیلوگرم بر متر مکعب، 71/1082 تومان بر مترمکعب، 03/528 تومان بر مترمکعب به دست آمد. متغیرهای بهره وری اقتصادی آب، ارزش فروش محصول (درآمد)، مقدار شکر تولیدی، هزینه های تولید، مصرف آب در هر دور آبیاری، فواصل آبیاری، هدایت الکتریکی خاک بعد از برداشت، هدایت الکتریکی آب رودخانه، متوسط زمان هر دور آبیاری، مساحت مزرعه، رقم گیاه، سن گیاه، تعداد دورهای آبیاری، pH آب خروجی زهکش و عرض خط کشت قبل از دیسکاور مهم ترین و تاثیرگذارترین متغیرها در الگوریتم های تصمیم CART و  CHAID می باشند. دقت مدل CART در قسمت آموزش و تست به ترتیب برابر 96 و 92 درصد بوده است و دقت مدل CHAID در بخش آموزش و تست به ترتیب برابر 97 و 90 درصد می باشد. همچنین بر اساس اولویت بندی عوامل موثر بر بهره وری با استفاده از روش تحلیل سلسله مراتبی به ترتیب کمیت و کیفیت آب آبیاری، شرایط آب و هوایی، شرایط گیاه، عوامل مدیریتی و انسانی و شرایط خاک با ضریب 459/0، 231/0، 150/0، 091/0 و 069/0 رتبه بندی شدند.

    کلیدواژگان: آب، بهره وری، تحلیل سلسله مراتبی، درخت تصمیم، نیشکر
  • رضا کریم زاده، حسین مقصودی*، حمیدرضا اخوان، کاظم جعفری نعیمی صفحات 113-123

    در پژوهش حاضر تاثیر تابش فرابنفش در محدوده UV-C با طول موج 254 نانومتر در بهبود ماندگاری دانه های انار ارزیابی گردید. برای این منظور ابتدا سامانه پیوسته پرتودهی فرابنفش ساخته شد و دانه های انار با دو دز 3/6 و 4/8 کیلوژول بر مترمربع پرتودهی شدند. نتایج نشان داد که اثرات ساده و متقابل تابش فرابنفش، نوع ظرف و مدت زمان نگهداری بر افت وزن و شاخص های رنگ معنی دار شد و به طور متوسط باعث 27% کاهش در افت وزن، %7 افزایش در شاخص L*، % 7/6 کاهش در مقدار a* و %10 افزایش در مقدار b* نمونه های کنترل نسبت به شاهد گردید. با افزایش دز پرتودهی شمارش کل باکتری ها و قارچ ها به صورت معنی دار و به میزان 65/1 چرخه لگاریتمی در هر گرم کاهش یافت. به علاوه، پرتودهی تاثیر معنی داری بر ویژگی های حسی مورد مطالعه (رنگ، طعم و مزه، عطر و بو، بافت و پذیرش کلی) داشت. به طور کلی، بر مبنای نتایج ارزیابی حسی، شاخص های رنگ و رشد میکروبی، دز تابش 3/6  کیلوژول بر مترمربع برای افزایش ماندگاری دانه های انار قرار گرفته در بسته بندی بدون منفذ پیشنهاد می گردد.

    کلیدواژگان: دانه انار، تابش فرابنفش، ویژگی های کیفی، ماندگاری
  • الهام بلندنظر، حسن صدرنیا*، عباس روحانی، مرتضی تاکی صفحات 125-137

    دما و کنترل آن در گلخانه یکی از پارامترهای مهم در گلخانه ها بوده و نقش اساسی در اقتصادی بودن تولید دارد. با وجود این که گلخانه یک محیط بسته است ولی کاملا از محیط بیرون ایزوله نیست. بنابراین شرایط داخل گلخانه تحت تاثیر تغییرات آب و هوایی بیرون دایما تمایل به تغییر دارد. هدف از اجرای این تحقیق، تخمین دمای هوا در یک گلخانه با پوشش پلی اتیلن با توجه به پارامتر های خارجی گلخانه شامل دمای هوا (Tout)، رطوبت نسبی هوا (Hout)، شدت تابش خورشید (S) و سرعت باد (V) با استفاده از روش های مختلف شبکه های عصبی مصنوعی شامل پرسپترون چند لایه (MLP)، تابع شعاع مدار (RBF) و عصبی-فازی (ANFIS) می باشد. مقایسه بین مدل های مختلف شبکه های عصبی نشان داد که روش RBF با ضریب تبیین بالاتر 93/0 = R2 و خطای کمتر 25/2 =RMSE نسبت به دو روش MLP و ANFIS دارای عملکرد بهتر در پیش بینی بود. نتایج ارزیابی مدل RBF برای پیش بینی دما در ساعات آینده بیانگر خطای قابل قبول در پیش بینی توسط این مدل تا دو ساعت آینده بود و بنابراین کشاورزان زمان کافی برای فراهم نمودن تمهیدات لازم جهت جلوگیری از افزایش دما در گلخانه در ساعات آینده و صرفه جویی در مصرف انرژی خواهند داشت.

    کلیدواژگان: دمای گلخانه، پوشش پلی اتیلن، شبکه های عصبی مصنوعی، مدل سازی
  • محمود صفری*، کریم گرامی صفحات 139-148

    در فرآیند تولید محصولات کشاورزی، انتخاب روش مناسب سم پاشی از اهمیت ویژه ای برخوردار است. از طرفی انتخاب مناسب ترین معیار برای ارزیابی و گزینش سم پاش ها از ضروریات است. در مرحله اول این تحقیق، با توجه به معیار های مهم ارزیابی عملیات سم پاشی، سم پاش های تراکتوری بوم دار، لانس دار، توربولاینر، میکرونر پشتی و اتومایزر در مناطق البرز، آذربایجان غربی، خوزستان و خراسان رضوی در قالب طرح آماری کاملا تصادفی به منظور سم پاشی مزارع گندم مورد مقایسه قرار گرفتند. در مرحله دوم، نتایج آزمایش اول به کمک روش تحلیل سلسله مراتبی (AHP[1])، تجزیه وتحلیل گردید. با توجه به نتایج به دست آمده، امتیاز روش های سم پاشی بر اساس معیارهای مصرف محلول سم، باد بردگی، ظرفیت مزرعه ای، یکنواختی پاشش، درصد لهیدگی محصول، توان اسمی سم پاش ها، اثربخشی و هزینه در نظر گرفته شد که پس از نرمال سازی و تجزیه وتحلیل داده ها، وزن سم پاش های میکرونر، بوم دار، توربولاینر، اتومایزر و لانس داربه ترتیب 337/0، 239/0، 175/0، 170/0 و 078/0 به دست آمد. مناسب ترین و ناکارآمدترین سم پاش در مزارع گندم، با در نظر گرفتن وزن های به دست آمده، به ترتیب سم پاش های میکرونر و لانس داربودند. ضریب ناسازگاری وزن ها، 08/0 بود. بالاترین و پایین ترین وزن معیار به ترتیب 253/0 و 038/0 و مربوط به اثربخشی در کنترل علف های هرز و آفات و لهیدگی محصول بود. در این شرایط ضریب ناسازگاری 09/0 به دست آمد.

    کلیدواژگان: الویت بندی، تحلیل سلسله مراتبی، سمپاشی
  • علی دینی، حسین قیومی زاده*، علی اکبر رحیمی فرد، علی فیاضی، محمدعلی افتخاری، مهدی عباس زاده صفحات 149-159
    هدف از این مطالعه توسعه الگوریتم های تصویربرداری، جهت بهبود درجه بندی آجیل ها با نقص های پوسته از جمله لکه های چربی، لکه های تیره، بدنه چسبیده، نقص های هسته ای آسیب و پوسیدگی قارچی می باشد. همه این نقص ها نشان دهنده خطر آلودگی به آفلاتوکسین هستند. شبکه های عصبی کانولوشن در زمینه های مختلف بینایی ماشین و طبقه بندی تصویر برجسته شده اند. در این پژوهش یک مدل سخت افزار آزمایشگاهی بر مبنای شبکه عصبی کانولوشنی جهت طبقه بندی پسته ها طراحی شده است. داده های جمع آوری شده 958 تصویر، شامل 276 تصویر از پسته های معیوب و 682 تصویر از پسته های سالم می باشند. طبقه بندی تصاویر شامل سالم و معیوب بر اساس 3 نوع شبکه عصبی کانولوشن عمیق شامل Google net، resnet18 و vgg16 انجام شده است. میزان دقت و ویژگی نتایج به دست آمده با استفاده از مدل های شبکه های عصبی عمیق از پیش آموزش داده شده Google net، resnet18 و vgg16 به ترتیب برابر 8/95%-1/97%، 2/97%-7/96% و 83/95%-08/97% هست.
    کلیدواژگان: آفلاتوکسین، پسته، شبکه عصبی عمیق، کانولوشن، طبقه بندی
  • سعید آقاعزیزی، منصور راسخ*، یوسف عباسپور گیلانده، محمدحسین کیانمهر صفحات 161-170
    یکی از مشکلات مهم مزارع تولید گندم وجود مواد خارجی همراه با محصول است. با جداسازی ناخالصی های توده گندم، ارزش اقتصادی گندم تولیدی و درجه خلوص بذر تولیدی افزایش می یابد. ازاین رو در این تحقیق از یک جداکننده میز وزنی برای جدا کردن ناخالصی ها از توده گندم استفاده شده است. دستگاه مذکور دارای پنج پارامتر قابل تنظیم سرعت هوا، دامنه نوسان، فرکانس نوسان، شیب طولی و شیب عرضی میز می باشد که تاثیر این پارامترها برای دستیابی به حداکثر جداسازی ناخالصی از توده گندم مورد بررسی قرار گرفت. آنالیز آماری در قالب دو آزمایش فاکتوریل در طرح پایه کاملا تصادفی انجام شد. در آزمایش اول اثر سه پارامتر شیب طولی، شیب عرضی و فرکانس نوسان میز و در آزمایش دوم اثر دو پارامتر دیگر بررسی شد. همچنین با استفاده از روش آنالیز ابعادی، پارامتر بدون بعد  حاصل شد که در بررسی اثر و کاهش تعداد پارامترها موثر بود. نتایج نشان داد در شیب طولی °5/2، شیب عرضی °5/1، فرکانس نوسان 395 سیکل بر دقیقه، دامنه نوسان 5 میلی متر و سرعت هوای 75/6 متر بر ثانیه بیشترین جداسازی ناخالصی ها از توده گندم برابر 03/87 درصد حاصل شد. همچنین با افزایش شیب طولی از °5/2 به°5/4، افزایش شیب عرضی از °75/0 به °25/2 (در اکثر موارد) و افزایش دامنه نوسان میز از 5 به 7 میلی متر جداسازی ناخالصی ها کاهش و با افزایش سرعت هوا تا 75/6 متر بر ثانیه جداسازی ناخالصی ها افزایش نشان داد.
    کلیدواژگان: گندم، جداکننده وزنی، ناخالصی، فرکانس نوسان
  • آیدین ایمانی، سلیمان حسین پور*، علیرضا کیهانی، مصطفی عظیم زاده صفحات 171-181
    توسعه هر نوع بیوسنسور با چالش هایی در زمینه بهینه سازی پارامترها و کالیبراسیون مواجه است. در این تحقیق رویکردی مبتنی بر یادگیری ماشین برای مدل سازی و بهینه سازی مولفه های تاثیرگذار در ساخت نانوبیوسنسور الکتروشیمیایی بر اساس الکترود کربن شیشه ای اصلاح شده با گرافن اکسید و نانومیله طلا در شرایط کاری آزمایشگاهی ارایه شده است. پاسخ نانوبیوسنسور به عنوان خروجی و تاثیر هشت عامل موثر شامل: غلظت گرافن اکسید، غلظت نانو میله های طلا، غلظت پروب تک رشته ای، مدت زمان ماند پروب تک رشته ای بر روی الکترود اصلاح شده، مدت زمان ماند MCH بر روی الکترود اصلاح شده، مدت زمان هیبریداسیون پروب و الیگونوکلیوتید هدف، غلظت محلول شناساگر اوراست بلو، مدت زمان ماند اوراست بلو، به عنوان ورودی های مدل شبکه عصبی برای آموزش و توسعه مدل مورد استفاده قرار گرفتند. نتایج بدست آمده نشان داد که خروجی مدل همخوانی قابل قبولی با نتایج آزمایشگاهی داشته و مدل می تواند پاسخ نانوبیوسنسور را با دقت 91/96 درصد و میانگین درصد خطای مطلق 5090/5 درصد پیش بینی کند. در پایان با استفاده از الگوریتم ژنتیک مقادیر بهینه متغیرهای ورودی برای دستیابی به حداکثر جریان پاسخ نانوبیوسنسور، محاسبه گردید. نتایج بهینه سازی نشان داد که این روش عملکرد مناسبی در مقایسه با نتایج آزمایشگاهی دارد و می تواند برای ساخت و طراحی نانوبیوسنسور مورد استفاده قرار بگیرد.
    کلیدواژگان: نانوبیوسنسور، مدل سازی، شبکه عصبی مصنوعی، بهینه سازی، الگوریتم ژنتیک
|
  • Abdollah Adibzadeh, Hassan Zaki Dizaji *, Nahid Aghili Nategh Pages 1-10

    Sugar cane is one of the most important industrial plants which is the first source of sugar production in Iran (about 40-50%). The sugar industry plays a key role among the various industries of the country with the daily supply of energy to the citizens. In addition to household consumption, sugar is of particular importance in the food industry because of its sweetening and volume properties. Sugarcane content is often in the range of 10-15% and in some cases up to 17%. Various factors such as variety and date of planting or harvesting of last year are important to start harvesting. On the other hand, Sugarcane cannot be stored in the factory and its sugars factors are decomposed quickly by storage and sugar cane weight decreases due to loss of moisture. It would be better if the factory consumes more fresh sugarcane. Therefore, an electronic nose instrument was used to test the variety of sugarcane syrup and its association with the odors emitted from it to identify the variety of sugarcane for harvesting time. Four sugarcane varieties (CP57, CP69, IRC99-02, and CP48) were selected from the sugarcane sample fields. Linear discriminant analysis (LDA), principal component analysis (PCA) and neural networks (ANN) were used to detect the different sugarcane varieties. The results showed that all three methods had high accuracy in variety classification. But the LDA and PCA methods performed better than the ANN method. So that, the classification accuracy of sugarcane varieties was 98.33%, 97% and 96.7%, respectively. The results showed the high ability of the olfactory machine to diagnose between the sugarcane varieties, which can be used as a rapid and low cost instrument in the sugarcane industry.

    Keywords: Sugarcane, Machine Olfaction, linear discriminant analysis, neural network, principal component analysis
  • Nahid Hasnaki, Yaghoob Mansoori *, Abbas Asakereh Pages 11-21
    The objective of the current study was to investigate the energy consumption in Karun Sugar Factory and to assess, economically and technically, the potential of substituting natural gas by bagasse as a source of energy. The necessary data were collected throughout the factory documents and the factory expert’s interview during years 1395-1396. The results showed that the annual consumption of electricity and heat in the Karun sugar factory was 24 GWh and 890.83 tonnes steam respectively, which resulted in the emission of 195 tonnes of air pollutants with a social cost of 24.79 billion Rials. The net heat rate of the factory’s power plant calculated to be 27.15 MJ/KWh. Using 270,000 tonnes of surplus bagasse, there is the potential to generate 15.14% of total sugar plant energy and reduce 29,000 tonnes of pollutants per year. The rate of return and return time of the capital needed for modification the current boilers was 88.95% and 2 years respectively. Using bagasse instead of natural gas leads to 30.6 Billion Rials reduction in variable costs of the factory based on financial data of the year 1395.
    Keywords: : Energy, Bagasse, Environmental, rate of return, Sugarcane
  • Sayed Javad Sajadi, Soleiman Hosseinpour *, Shahin Rafiee Pages 23-35
    Aflatoxin B1 (AFB1) is one of the most toxic Aflatoxins that contaminates agricultural products and causes deathlike effects on human health. Determination of AFB1 in food by biosensors is fast, low cost and accurate. In this paper, modeling and simulation of chemical reactions in the AFB1 potentiometric biosensor is performed to determine the optimal reaction rate constants. Enzymatic reactions are simulated using COMSOL software and reaction rates are optimized by Artificial Neural Network (ANN) and Genetic Algorithm (GA). The fitness function of GA is defined by deploying ANN. The data generated during the simulation step were used to train and evaluate the performance of the neural network. Compared with experimental data, COMSOL model simulated biosensor response with MAPE equal to 0.1023 %. In addition trained ANN with 5-48-1 structure predicted biosensor response with MAPEs equal to 0.7074 %, 0.9458 %, 0.7473 % and 0.7492 % for train, validation, test and total data sets respectively. Reaction rates were optimized by Artificial Neural Network (ANN) and Genetic Algorithm. Modeling results showed that trained Neural Network using Genetic Algorithm optimized reaction rates has the lowest MAPE equal to 0.0026 % compared with other models in prediction of AChE enzyme inhibition by AFB1.
    Keywords: Biosensor, modeling, COMSOL, Artificial Neural Network, Genetic Algorithm
  • Reza Torkashvand, Reza Amiri Chaijan *, Ali Ghasemi Pages 37-49

    In this study, the effect of effective variables of moisture content, particle size, type of adhesive and tablet shape on the qualitative properties (difference of color indices ΔL*, Δa* and Δb*), physical (unit density and shrinkage) and mechanical (diffusion force) indices of compressed tablets made from tomato powder were studied. Independent variables were including three levels of particle size of tomato powder (particles smaller than 0.3 mm, particles ranging from 0.3 to 0.75 mm and particles larger than 0.75 mm), three levels of moisture content (18, 36 and d.b. 54%), three forms of compressed pills (spherical, cylindrical and cubic), three types of adhesive (55% fructose syrup, water and sugar). For optimization, the surface response and Di-optimal method were used. The results showed that optimum spot was obtained for compressed pills consisting of tomato powder with a moisture content of %33.4 db, particle size of fructose 0.3 mm, and cylindrical tablet form, dried in oven at 60 ° C. Under these conditions, the desirability index was 0.826, and the optimal value of the independent variables ΔL*, Δa* and Δb* (the difference of color indices with fresh tomatoes), penetration force, unit density and shrinkage were 35.89, 15.6, 23.23, 267.2 N, 2299 kg/m3 and %12.2, respectively.

    Keywords: Optimization, Compressed tablets tomato, Moisture content, Particle size
  • Mohammad Ahmand, Faroogh Sharifian *, Ali Mohammad Nikbakht, Vahid Rostampour, Edris Rahmati Pages 51-61
    In this paper the energy and exergy analysis of flat plate solar collector simulator equipped with Inclined Broken Rib roughness was investigated based on experimental data in open circuit as well as optimizing system operating conditions. The experiments were carried out with nine levels of mass flow rates (0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, 0.10 and 0.11 kg/s), five levels of heat flux (1000, 1100, 1200, 1300 and 1400 W/m^2) and three levels of ambient air temperature (20, 25 and 30 °C). The results showed that the highest and lowest values of energy efficiency were found 49.8 and 0.3%, in treatments with ambient temperature of 20 and 25 °C, mass flow rate of 0.11 kg/s and heat flux of 1000 and 1400 W/m^2, respectively. Also, the highest and lowest exergy efficiency were calculated 5.75 and 0.607% in treatments with ambient temperature of 20 and 30 °C, mass flow rate of 0.11 kg/s and 0.03 and heat flux of 1000 W/m^2, respectively. The response surface methodology was employed to optimize solar collector operating conditions. Optimum operating conditions were found to be anambient temperature of 20 °C, mass flow rate of 0.11 kg/s and heat flux of 1000 W/m^2. At this optimum condition, the energy and exergy efficiency were found to be 42.08 and 5.76%, respectively at a desirability level of 0.92.
    Keywords: Exergy efficiency, Energy efficiency, Solar Energy, Response Surface Methodology
  • AhmadReza Salehiyoun, Mohammad Sharifi *, Mortaza Aghbashlo, Hamid Zilouei, Saeed Mofatteh Pages 63-76

    Anaerobic digestion in order to produce biogas is a proven method for producing renewable energy from municipal solid waste. In this research, organic fractions of municipal solid waste compounds were determined in Karaj metropolitan area. The organic waste components were monitored in five categories of fruits, fat and protein, starches, vegetables, and cellulose wastes in the winter and summer seasons. Then, a sample representing the average amount of waste components was synthesized and biomethane yield, digestibility indicators and kinetic modeling parameters of biogas production were investigated in batch tests at mesophilic temperature at two concentrations of 8 and 15 TS%. The most part in the organic fraction was fruit and vegetable waste with a total of 62.9% and 70.6% in winter and summer, respectively. The biomethane yield and methane content at 8% and 15% TS had significant difference with 385.2 and 289.2 L/kg VS and of 66.8 and 58.8%, respectively, but there was no significant difference for VS removal with 87.99% and 84.72%. As a result, for source separated MSW, anaerobic digestion at the lower TSs has better results than dry. Continuous anaerobic digestion at 30 day hydraulic retention time is more effective for specific biomethane production and high volumetric biogas production under stable conditions.

    Keywords: Anaerobic digestion, Municipal solid waste, Biogas, Batch test, modeling
  • Mohammad Sharifi *, Leyla Behrooznia, Seyed Hashem Mousavi Avval Pages 77-87

    Municipal solid waste, known as reject material, is increasingly being added to the world, causing more problems, such as air pollution and greenhouse gas emissions in the environment. As a result, the need for proper and sustainable waste management is felt more by managers. Accordingly, compost production is one of the methods used in agriculture in addition to reducing pollution. In Rasht, this method is used to manage 400 tonns of waste per day. In the process of composting, pollutants are created that affect the environment. In this study, the CML-IA baseline V3.04 / World 2000 method using SimaPro software was applied to evaluate the life cycle and 11 impact categories have been investigated and finally the results were normalized and weighed. The functional unit in this study was 400 tonnes of compost produced per day. The results showed that the global warming potential was calculated with 4.28×103 kgCO2 and the largest share in this section was due to direct emissions and transportation. Also, normalization results showed that compost production from waste had the most effect on marine aquatic ecotoxicity and human toxicity potential, respectively.

    Keywords: Global warming potential, greenhouse gas emissions, Environmental Pollution, compost
  • Kamran Maleki Majd, Dariush Zare *, Emdad Homayoun, SAYED MEHDI NASIRI, Gholamreza Karimi, KHOSROW JAFARPUR Pages 89-97
    In this study pressure drop of bin containing corn as granular porous material with two common inlet duct arrangements (H, F) for a laboratory silage was studied by using experimental method. Higher pressure was obtained for the F arrangement compare to the H arrangement. The Ergun equation was fitted on the pressure drop of empirical data and the correction coefficients of modified Ergun equation for both H and F arrangements were determined. The coefficient of determination, mean root square error and mean relative error were used as indices to investigate the goodness of fitness between the empirical data and Ergun equation, and the equation had good fitness on the empirical data. The numerical simulation was carried out by finite element simulation of COMSOL Multiphysics v5.3 software using the modified Ergun equation. Moreover, distribution of velocity and flow lines within the bin was also presented.
    Keywords: Bin Aeration, Numerical simulation, Granular Material, Duct inlet arrangement
  • Nasim Monjezi * Pages 99-112

    The purpose of this study was to identify, prioritize and control the factors affecting water productivity in sugar cane production. In this study, to calculate water productivity, the indices such as BPD and CPD, NBPD is used. Given that the research goal was to prioritize the factors affecting the productivity of water input during the sugar cane production process, the identified factors were evaluated using Analytical Hierarchy Process Analysis (AHP). Then, CART and CHAID decision trees were used in modeling the water input efficiency and the factors influencing it. According to the results, the average CPD, BPD and NBPD indices for sugar cane were 2.37 kg m-3, 1082.71 toman m-3, and 528.03 toman m-3, respectively. The variables of economic productivity of water, the value of sales of the product (income), the amount of sugar produced, production costs, water consumption per irrigation interval, irrigation intervals, electrical conductivity of the soil after harvest, river electrical conductivity, mean time of each irrigation interval, Plant height, plant age, number of irrigation cycles, drainage water drainage PH and width of the cultivating line before discovery are the most important and influential variables in the decision tree models of CHAR and CHAID. The accuracy of the CART model in training and testing was 96% and 92%, respectively, and the accuracy of the CHAID model in education and testing is 97% and 90%, respectively. Also, based on the prioritization of the factors affecting productivity, using the hierarchical analysis method, the quantity and quality of irrigation water, climate conditions, plant conditions, managerial and human factors, and soil conditions were ranked with a coefficient of 0.459, 0.231, 0, 0.091 and 0.069, respectively.

    Keywords: Analytical Hierarchy process, Decision Tree, productivity, Water, Sugarcane
  • Reza Karimzadeh, Hossein Maghsoudi *, HamidReza Akhavan, Kazem Jafari Naeimi Pages 113-123

    In this present study, the effect of ultraviolet radiation in the UV-C range with the wavelength of 254 nm on the increasing shelf-life of pomegranate arils was evaluated. For this purpose, at the first, a continuous ultraviolet irradiation system was constructed and pomegranate arils were irradiated with two doses of 6.3 and 8.4 kJ/m2. The results showed that the simple and interactive effects of UV irradiation, container type and storage time on the weight loss and color indices were significant and on average caused 27% decrease in weight loss, 7% increase in the L* index, 6.7% decrease in the a* value and 10% increase in the b* value of the control samples. With increasing irradiation dose, the total bacterial and fungal count were significantly reduced by 1.65 Log cfu g-1. Furthermore, the irradiation had a significant effect on the studied sensory properties (aroma, color, texture, and overall acceptance). Generally, based on the results of sensory evaluation, color indices and microbial growth, irradiation dose of 6.3 kJ/m2 is recommended to increase pomegranate arils shelf-life in non-porous packaging.

    Keywords: Pomogrenate arils, Ultraviolet irradiation, Quality properties, shelf-life
  • Elham Bolandnazar, Hassan Sadrnia *, Abbas Rohani, Morteza Taki Pages 125-137

    Internal temperatures of greenhouse and its control is one of the important parameters in greenhouses and plays a key role in the economics of production. Although the greenhouse is a closed environment, it is not completely isolated from the outside. Therefore, the conditions inside the greenhouse are constantly changing under the influence of outside climate change. The purpose of this study was to estimate the internal air temperature of polyethylene greenhouse with respect to the external parameters of the greenhouse including air temperature (Tout), air relative humidity (Hout), solar radiation (S) and wind speed (V). For this purpose, different method of artificial neural networks including Multilayer Perceptron (MLP), Radial Basis Function (RBF) and Adaptive Nero Fuzzy Inference System (ANFIS) were used. Comparison between different neural network models showed that RBF method had better prediction performance than MLP and ANFIS with higher coefficient of determination (R2=0.93) and lower error (RMSE=2.25). The results of the RBF model estimation for the prediction future temperature indicated an acceptable error in the prediction by the model for the next two hours and thus, the farmers had enough time to provide the necessary measures to prevent the greenhouse temperature rise in the future and save in energy consumption.

    Keywords: Greenhouse temperature, Polyethylene Cover, artificial neural network, modeling
  • Mahmood Safari *, Karim Gerami Pages 139-148

    In the process of crop production, selecting the appropriate spraying method is very important. On the other hand, selecting the most appropriate criterion for evaluating and selecting pesticides is essential. In the first stage of this research, according to the important criteria of evaluation of spraying operations; the sprayers of Knapsack micronair, Tractor boom, Tractor lance, Knapsack atomizer and Turboliner were compared for sparying wheat fields in Alborz, West azarbaijan, Khoozestan and Razavi khorasan proviences with a completely randomized design. In the second step, the results of the first experiment were analyzed using AHP method. According to the results, the scoring methods were evaluated based on the criteria of spraying, spraying volume, drift, field capacity, spray uniformity, crop blunder, spraying nominal power, effectiveness and cost, after normalizing and analyzing the data, weight of Knapsack micronir, Tractor boom, Turboliner, Knapsack atomizer, and Tractor Lance sprayers were 0.337, 0.223, 0.175, 0.0 170 and 0.078, respectively. The Knapsack micronair and Tractor lance sprayers was the best and worst sprayer according to the weights obtained. The inconsistency coefficient of weights was 0.08. The highest and lowest criterion weight was 0.253 and 0.038 respectively and related to the effectiveness of controlling weeds and pests and crop blunder. In this situation, the inconsistency coefficient was 0.09.

    Keywords: Analytical Hierarchy Process (AHP), spraying, Sprayer
  • Ali Dini, Hossein Ghayoumi Zadeh *, Aliakbar Rahimifard, Ali Fayazi, Mohammadali Eftekhari, Mehdi Abbaszadeh Pages 149-159
    The aim of this study is to develop imaging algorithms to improve the grade of nuts with shell defects such as oily stains, dark stains, adhering hull, damage seed defects, and fungal decay. All these defects indicate the risk of Aflatoxin contamination. Convolutional Neural Networks (CNNs) have become prominent in various fields of machine vision and image classification. In this study, a laboratory hardware setup based on a convolutional neural network is designed for sorting pistachios. The total number of collected data is 958 images, which includes 276 images of defective pistachios and 682 images of healthy pistachios. The classification of healthy and defective images has been accomplished by 3 types of deep convolutional neural networks including Google net, resnet18 and vgg16. The accuracy and specificity of the results obtained using the pre-trained deep neural network models of Google net, resnet18 and vgg16 are 95.8% -97.1%, 97.2% -96.7%, and 95.83% -97.08%, respectively.
    Keywords: Aflatoxin, Pistachio, Deep Neural Network, Convolution, Classification
  • Saeed Agaazizi, Mansour Rasekh *, Yousef Abbaspour Gilandeh, Mohamadhosein Kianmehr Pages 161-170
    Presence of foreign materials with the product is one of the important problems of wheat production. The economic value of the produced wheat and the degree of purity of the produced seeds increases with the separation of wheat mass impurities. Hence, in this research, a gravity separator table was used to remove impurities from wheat bulk. The machine has adjusting five parameters of air velocity, frequency of oscillation, amplitude of oscillation, longitudinal slope and latitudinal slope of the table. The effect of these parameters was studied to achieve maximum impurity separation from wheat bulk. Statistical analysis was performed in two factorial experiments based on completely randomized design. In the first experiment, the effects of three parameters of longitudinal slope, latitudinal slope and frequency of oscillation of the table were investigated and in the second experiment the effect of two other parameters was investigated. Also, using dimensional analysis, a dimensionless number parameter was obtained which was effective in evaluating the effect and reducing the number of parameters. The results showed that the maximum separation of impurities from wheat bulk was 87.03% at longitudinal slope of 2.5 °, latitudinal slope of 1.5 °, frequency of oscillation of 395 cycles per minute, amplitude of oscillation of 5 mm and air velocity of 6.75 m/s,. Also, with increasing longitudinal slope from 2.5 ° to 4.5 °, latitudinal slope from 0.75 °  to 2.5 ° (in most cases) and the amplitude of oscillation of the table from 5 to 7 mm, the separation of impurities was reduced and with increasing the air velocity from 5.25 to 6.75 m/s the separation of impurities was increased.
    Keywords: Wheat, Gravity separator, Impurity, Oscillation Frequency
  • Aydin Imani, Soleiman Hosseinpour *, Alireza Keyhani, Mostafa Azimzadeh Pages 171-181
    Developing a biosensor faces the different challenges for parameter optimization and calibration. In this study, a machine learning based approach is used to model and optimize the effective parameters of an electrochemical nanobiosensor based on thiolated probe-functionalized gold nanorods (GNRs) decorated on the graphene oxide (GO) sheet on the surface of a glassy carbon electrode (GCE). The response of the biosensor was considered as the output and eight effective factors including GO concentration, GNR concentration, probe concentration, probe time, MCH time, hybridization time, Oracet Blue (OB) concentration, and OB incubation time were used as inputs to train and model an artificial neural network. The experimental results demonstrate that the output of the developed model has an acceptable compatibility with the results obtained in the laboratory. The developed model is able to predict the output of the nanobiosensor with accuracy of 96.91% and the mean absolute percentage error (MAPE) value of 5.5090 %. Finally, genetic algorithm is used to find the optimum values of these parameters which yield the maximum value of the nanobiosensor output. The optimization results indicated that this method has better performance compared to the laboratory results and this method can be used for nanobiosensor design.
    Keywords: Nanobiosensor, modeling, Artificial Neural Network, Optimization, Genetic Algorithm