Design of Pumping Machine Servo Motor Controller Based on Fuzzy Control
According to statistics, at present, the number of pumping units in use in China is more than 200,000 units, and it is increasing at an annual rate of about 15,000 units. The average power of motors used in "pumping units" is around 40KW, the vast majority. The pumping unit still adopts the beam-type structure. This structural feature is simple and reliable, but the efficiency and power factor of the whole machine are relatively low, and the energy consumption is also large. The annual power consumption is more than 15 billion degrees, so the petroleum industry Large-capacity households are also energy-intensive households. China is a poor oil country, mainly replacing oil with water and oil for electricity. The electricity consumption of pumping units accounts for 30% to 45% of the total cost of oil production. Therefore, the development of energy-saving controllers for pumping units is very high. Pay attention to it, so the new energy-saving pumping equipment that reduces the energy consumption and improves the efficiency of the pumping unit will become the development trend and target of the pumping unit production industry in the future.
At present, most of the domestic long-stroke pumping units use switched reluctance motors. Because the switched reluctance motor has large magnetic circuit saturation, the double salient pole structure and the switch control mode lead to its high nonlinearity. However, due to the serious nonlinearity of the switched reluctance motor and the characteristics of variable parameters and variable structure, it is difficult to achieve the ideal control performance by using the PID controller with conventional fixed parameters. The control parameters cannot be accurately established because of its mathematical model. It's hard to get ok. In order to adapt to the nonlinear characteristics of switched reluctance motors, a fuzzy control strategy with variable parameters is adopted. Combining artificial neural network with fuzzy control, the neural network's adaptability, self-learning ability and nonlinear mapping ability are fully utilized to form a fuzzy neural network control strategy with strong adaptive parameters.
Fuzzy neural network (Fuzzynetwork-FNN) is a combination of fuzzy control theory and neural network control theory. It contains many advantages of fuzzy theory and neural network. It is a combination of learning, association, recognition and information processing. Fuzzy control is an intelligent control method that is widely used in engineering fields. It mainly transforms the experience of manual control into a control strategy. Therefore, it is not necessary to establish an accurate mathematical model of the controlled object, and its dynamic quality is better than ordinary. Control method. However, since the ordinary fuzzy controller is essentially a PD regulator, the static performance is not good, and there is a static difference. In order to solve this problem, this paper proposes a control method combining fuzzy control and neural network control, and adds the integral. The link is used to overcome static static differences.
