Aiming at the characteristics of complex control structure, variable working conditions, large inertia and strong coupling of boiler combustion system in thermal power plant, a control method of boiler combustion system based on Improved Fuzzy PID controller is proposed. Particle swarm optimization (PSO) is used to optimize the adjusting factor of the fuzzy controller on-line, which overcomes the control difficulties of the combustion control system due to the mathematical model of variable working conditions.
The simulation results of conventional PID controller and fuzzy adaptive PID controller show that the improved fuzzy PID controller has stable output, high control precision and good robustness.
The production process of large thermal power unit can be divided into boiler combustion system and steam-water system. Boiler combustion system is an important control system to provide heat to maintain the main steam load and ensure combustion economy and safety.
Main steam pressure is an important indicator to measure whether the steam volume and external load are compatible. Due to the large capacity of large units, the variety of fuel, and the number of coal mill feeders put into operation, it is difficult for conventional PID controller to meet the requirements of real-time control. Fuzzy controller is a simple non-linear controller with good robustness, adaptability and fault-tolerance. Some scholars have applied it to thermal control system of thermal power plant. However, due to the fact that fuzzy control is essentially a kind of non-linear PD control, the static error of the system can not be eliminated. The core idea of the improved fuzzy control strategy is to adjust the fuzzy universe and membership function according to the values of input error E and error change EC while keeping the fuzzy segmentation unchanged in the fuzzy universe. Fuzzy adaptive PID controller is a composite controller of fuzzy control and PID control. The controller changes the parameters Kp, Ki and Kd of traditional PID controller, which can not adjust the error in real time.
A method of using fuzzy controller to track the error signal to change the parameters of PID controller online is proposed, which improves the effect of fuzzy control. Particle swarm optimization (PSO) is a bionic optimization algorithm. In this paper, PSO is used to search and optimize the scaling factor, and then improve the control effect of the improved PID controller. The specific optimization process is as follows: the parameters to be optimized are alpha e, alpha ec, beta P and beta i, which constitute four dimensions of the search space. They randomly generate a set of Xi. As the first generation initial population, Xi is brought into the objective function Q to calculate the fitness value. 。 Repeat the above steps until the optimal solution is obtained. Fuzzy adaptive PID controller can modify the control parameters of the original PID controller, but the control accuracy is limited. In this paper, the standard particle swarm optimization (SPSO) algorithm is used to optimize the adjusting factors of the fuzzy adaptive PID controller online. Combining the advantages of the two controllers, the universe and output of the fuzzy controller can be adjusted in real time according to the size of the system error, so as to improve the control accuracy of the system.
Experiments were carried out on the combustion system process picture of a 300 MW unit in a power plant. Sampling period was 5 seconds and total sampling time was 20 minutes when different initial conditions were loaded. After the data are processed, the mathematical model of the fuel control system under different loads is obtained by using the identification algorithm in Matlab. This paper chooses the improved PSO algorithm to optimize the adjustment factors of the input and output links of the aforementioned fuzzy adaptive PID controller. The fuzzy word set of the fuzzy controller is {PB, PM, PS, Z, NS, NM, NB} and the basic universe of the input variables E and EC is set to [-12, 12], [-6, 6].
The basic universe of the output variables should be set according to the parameters of the PID controller. Through the parameter tuning method of the PID controller, the parameters of the PID controller are obtained as follows: delta = 0.
48, Ti = 289, Td = 0.0001.
Therefore, the basic universe of the output of the fuzzy controller should be selected in a certain range of proportion and integral coefficients. For the convenience of calculation, it should be set as [-0.6, 0.6], [-0.012, 0.012], while the fuzzy universe of the corresponding fuzzy variables of the input and output should be [-6, thermostatic element 6].
The method of maximum membership degree is used to de-fuzzify. According to formula (10), the quantization factors Ke and KEC of input variables are 0.5 and 1 respectively, and the ratio factors KP and Ki of output variables are 0.1 and 0.
001 respectively. In order to verify the control effect, compared with the conventional PID controller and the fuzzy adaptive PID controller, the output of the system with three control modes under constant jump disturbance is simulated. By comparing the three curves, it can be seen that the improved fuzzy-PID controller is better than the other two controllers in dynamic and steady-state performance. Its control precision is higher and the curve is more stable. In the fuel control system, the model parameters are selected when the unit load is 80% and 100% respectively, and the robustness of the two controllers is observed by adding a constant disturbance of amplitude 1 while keeping the initial values of PID, control rules and particle swarm optimization of the fuzzy adaptive PID controller and the improved controller unchanged. The robustness of the improved controller is obviously better than that of the fuzzy adaptive PID controller. In this paper, an improved fuzzy-PID controller is designed based on the idea of variable universe control and particle swarm optimization. Standard particle swarm optimization (SPSO) is used to optimize the adjusting factor of the fuzzy controller on-line and change the output of the controller, thus reducing the dependence of the control system on the fuzzy control rules and the control system model, and optimizing the control effect.