The Variable-frequency Energy-saving Control System Of The Pumping Unit Based On Intelligent Control

The beam-pumping unit occupies an important place in the oil production industry. It will continue to play a major role in oil production and serve as the major energy consumption equipment for quite a long time for its simple structure, high reliability, fine durability, easy maintainance and adaptation to site condition. The installed motor power is often larger than than that required in normal conditon to guarantee adequate overload capacity. But the load is low and the motor load rate is only about 30% during most of the oil production process, especially in the late stage of oil production. The energy waste of oilfield due to pumping unit is quite large, and energy-saving measures should be taken.Aiming at problems such as light load and "dry tap", this paper presents a synthesis energy-saving scheme with the combination of intelligence control and variable frequency speed regulation. In this shceme, fuzzy neural network is used to predict the storage volume of underground oil. It is designed to turn off the motor in time when the volume is small and start the motor timely when the volume meets the "full tap" condition. When starting the motor, the variable frequency speed regulation technology is used to make a real-time controll of the speed according to the load size to achieve better objective of energy-saving.In order to realize the energy-saving scheme, technologies of intelligence control and variable frequency speed regulation are studied. An intelligent controller an a frequency converter are designed according to the characteristics of pumping unit, and the interface between them are reserved for the sake of interconnection.In designing the intelligence control, this paper analyses the characteristics of fuzzy control and neural network, respectively. The fuzzy control has fine ability of expression, and the neural network has better ability of study. In order to take the full advantage of the both, the TS fuzzy system model, combination of fuzzy control and neural network, is chosen to realize the intelligence control. According to the characteristics of oilfield, a small region of oilfield is chosen as the control object. A fuzzy neural network of the beam-pumping unit control is designed. To obtain the practical fuzzy rules and training network, we went to the Shengli Oil Field Design Institute and oil production plant, and got lots of real data from the staff and experiments done with some selected light load oil well. All these data is used in network training. Fuzzy rules are extracted from some oil experts. Part of the hardware and software of the intelligence controller are designed.Since the general frequency convertors are very expensive, its application in the oil field is not good. In this paper, we designed a new frequency converter for beam-pumping unit with the engineers. The dedicated frequency convertor consists of Single-Chip system, rectifier/inverter and IGBT driving-protection circuit. This convertor is simple but practical and costs little. The main circuit of the frequency conversion system is improved in order to meet the requirement of frequency speed control.In addition, fictitious load of bumping uint is designed to simulate the oil production process. It is made up with induction motor, gearbox, frame and weights. A lot of experiments are made on this fictitous load with the energy-saving control system. In the end, the experimental results are analyzed and the performances of the proposed energy-saving system is tested.