fuzzy controller

This paper presents a new method for using data collected from the agent's random movement in the environment for the initial adjustment of parameters of a controller with a fuzzy reinforcement learning structure. Slow learning speed and high failure rates during training are two major challenges in such structures. The initial parameterization of the fuzzy system can be a suitable solution to address these challenges. In this paper, the method of discrete value iteration is extended to continuous without relying on derivative based methods to initialize the parameters of the fuzzy system. First, random interaction with the environment is used to collect relevant data. Since the state space is continuous, the data is appropriately clustered and each cluster is considered as a state. Then, by generalizing the standard value iteration method to the continuous, the transition probability matrix and the immediate reward expectation matrix are calculated. Using the results of this stage, the initial parameterization of the fuzzy reinforcement learning structure is performed. Subsequently, these parameters are fine-tuned using reinforcement learning. The proposed method is called "Value Iteration based Fuzzy Reinforcement Learning" and is used in the problem of target following robots. The experimental results indicate a significant improvement in the performance of the proposed method in the problem of target following robots.

In the present study, a low-power power function generator circuit is proposed for fuzzy applications. The proposed power function generator circuit consists of squaring, square root, and analog multiplier circuits. All the circuits are designed in the subthreshold region to achieve minimum power consumption. The proposed power generator module is based on a fuzzifier circuit, and the analog multiplier circuits are used to adjust the slopes of the fuzzy functions. Besides performing the mentioned adjustments, the proposed circuit can adjust the rising and falling slopes quite separately. Analog multipliers are used in the power generator part to generate desired powers continuously with minimum number of control inputs. The proposed structure is presented in 0.35 μm technology, and the simulation results show that at a supply voltage of 1.3 V, the values of power consumption and error are respectively equal to 0.0036 μw and 0.8%, indicating the improvement of the proposed structure in terms of error and power consumption compared to the best relevant structures in the literature.

Nowadays, the use of renewable energy has become very important. Considering the changes in radiation intensity and the continuous switching of the inverter, the control algorithm of the inverter plays an important role. In the DC part, adjusting the voltage of the DC link is very important, and stabilizing the voltage at this point will have a significant impact on the performance of the entire system. In the existing methods in order to regulate the voltage, a proportional integral controller (PI) is used and the coefficients of this controller are usually fixed and its performance decreases when the working conditions change, so using structures with an approach Fuzzy is noticeable. The use of the fuzzy PI controller improves the performance of the controller, increases the speed of realizing the maximum power tracking algorithm, reduce the fluctuations of the DC link voltage and the output power of the photovoltaic system inverter. In this article, a new structure based on the fuzzy PI controller is presented in order to improve the performance of the inverter control algorithm, so that the DC link voltage becomes equal to the reference value for 2 seconds faster than the normal PI method and the grid voltage range for 1/5 seconds faster. And the active power of the grid will reach a steady state 2 seconds faster and the fluctuation of these variables will be significantly reduced.

Electricity is one of the most important carriers of energy used in buildings. By introducing energy smart grids (SG) and energy micro smart grids (MSGs) alongside smart buildings, a good platform has been provided to optimize electricity production and consumption. In this paper, an MSG consists of renewable resources, diesel generators, and cell batteries in bidirectional connection with an SG is modeled, which is connected to a smart building, including lighting, temperature, and air conditioning loads on the consumer side Fuzzy controllers have been used to implement intelligent building management systems and MSG to manage power generation and sales with SG. The decision variables are fuzzy controller rules optimized by an NSGAII multi-objective optimization algorithm with five different objective functions. The results show that the fuzzy controllers work better than the methods directly using the power allocated to the building loads as the decision variable.

Acquiring appropriate tools adaptable to various tasks is the most fundamental feature of the robots dealing with environment. Therefore, it would be more beneficial to plan a mechanism amenable to control the robotic fingers. Since robotic fingers have completely nonlinear behavior and their modeling is associated with the difficulties arising from the factors such as friction, physical features of transmission mechanisms, and changes in hand’s orientation, adopting a model-independent method of control will be useful. In this paper, a Takagi-Sugeno-Kang (TSK) fuzzy controller, which adaptively updates its consequence parameters, is employed for position/force control of the fingertips grasping a light and soft ball. Designing in Cartesian space and being model-independent are some of the most important advantages of this method. In the first step, force and position reference values are calculated using a predetermined stationary grasping strategy. Afterward, the performance of the adaptive fuzzy TSK controller in maintaining the ball with and without the existence of force measurement noise and joint friction are evaluated. Furthermore, the process of catching a falling ball is divided into approaching, locking and holding phases. Finally, in the simulation section, it is shown that the adaptive fuzzy TSK controller is an efficient way for performing the aforementioned three phases.

The P-V characteristic of a photovoltaic system is a nonlinear characteristic. The location and the value of the maximum power point in a Photovoltaic (PV) system depends on environmental conditions such as the intensity of solar radiation and the ambient temperature. To extract maximum power, a system including a DC-DC converter and a control method is required. Such a system is called maximum power tracking (MPPT), which is an essential part of a photovoltaic system. Model predictive control (MPC) is a type of control method which is used in MPPT. The model-based predictive control method detects the maximum power point using the equations of the dc-dc converter. In MPC, the system equations are ideally formulated. However, the values of the elements in the dc -dc converter change over time. There are two solutions to this problem. The first is to write real system equations. As a result, the formulations become very complicated and take up a large amount of microprocessor time and memory. Also, most of the time the exact values of the elements in the circuit are unknown. The second method is the use of additional sensors, which increases the cost of the system. In this article, a Fuzzy Model Predictive Control (FMPC) method is proposed for a Cuk converter to implement MPPT in photovoltaic systems. The proposed method obtains an improved performance and eliminates the need for the actual values of the elements. The method has been simulated in the MATLAB/Simulink workspace and its performance superiority has been demonstrated by comparing it against existing MPC and FL- MPPT approaches.