The problem of supply and discharge of sucker rod pumping system is the fundamental problem that affects system efficiency and well potential. However, with the current common technology and equipment level, it is very difficult to solve all aspects from design to on-site management. Pole pump intelligent control system to address the issue, the integration of information, automation, intelligent technology, established a dynamic surface echo identification model and calculation methods and supporting calculation control software, including the oil wells dynamic surface, , Including real-time monitoring of operating parameters; the pumping strength real-time control, thus solving the pumping system scheduling problems. After field test, this technology has achieved the expected goal, and has good effect in improving system efficiency, lengthening inspection pump cycle, saving electricity and so on. It is a new development of optimization technology of rod pumping system. There are more than 100,000 oil wells in our country. Among them, 80% of oil wells adopt sucker rod pumping technology. The supply and discharge problems of sucker rod pumping system are the fundamental problems that affect system efficiency and well potential. However, with the current common technology and equipment level , From design to site management are very difficult to solve in all aspects. The optimization design includes two types of content: First, the optimal design of new oil wells, that is, according to the geological department's pre-production and long-term development needs of the oilfield, the reasonable choice of the best machine, rod, pump combination, and determine the pumping parameters; Old wells pumping parameters optimization [1]. In the design process, oil well distribution is an important constraint, due to the pumping parameters is not continuous, it is difficult to achieve the design displacement is equal to the oil well with the production, mining applications generally make the design displacement slightly larger than the oil well production . At the same time, the allocation of new oil wells is also based on the expectations of all aspects of reservoir conditions, with the actual production of oil wells there is a certain gap, and sometimes may be great. The existence of these two factors makes the pumping system in the design stage has the possibility of not adjusting the supply and discharge. With the continuous development of reservoir development, oil well productivity is affected by many aspects, such as geological characteristics, reservoir management, oil production engineering, production and maintenance, etc., and it is a dynamic variable from a long period of time. When the amount of oil supply changes, it is necessary for the old well pumping parameters optimization, otherwise, when the lack of liquid, there will be empty pumping phenomenon, so that energy consumption increased. When the fluid is sufficient, pumping strength is low, it will limit the amount of liquid production wells. The dynamic liquid surface parameters of oil well directly reflect the fluid supply conditions of the formation and the relation between the supply and discharge in the well, which is one of the key data for evaluating and optimizing the adaptability of oil recovery technology. At present, subject to various factors such as automation level and personnel cost, the conventional practice of oilfields in China is to test the dynamometer, the liquid level and other technical data on a monthly basis to determine the theoretical operating parameters (such as: stroke, speed, row Volume, etc.), due to data acquisition, analysis, adjustment of the stage, so that the optimal adjustment showed after control, lost the best time; due to the limitations of the existing technology and equipment, regardless of stroke or stroke adjustment are not continuous, it is difficult to achieve Theory requirements; at the same time by the human factors, often in the actual work is not timely reference, the operating parameters of the phenomenon of unreasonable, both labor-consuming and can not achieve the best results. In short, the current technical means is difficult to achieve rod pump system supply and adjustment. For a long time, this issue has been widely concerned by the petroleum community. Rod pump intelligent control system is developed for the above problems. Intelligent control technology includes two real-time concepts: ①real-time monitoring, that is, the ability to collect downhole data and / or dynamometer; ② real-time control, that is, through the automatic actuator control pumping capacity. The system from the micro-point of view of the time to achieve scheduling. First, the composition and working principle 1, real-time monitoring of real-time monitoring of wells can be part of the liquid level, Kung-fu map and other operating parameters. The movable liquid level detecting device consists of an infrasonic wave sending device, an infrasonic wave receiving device and a data collecting interface. The infrasound generating device adopts a spring-loaded aeroelastic device as a sound source, and at a set time, accumulates it with a compression spring of a reducer and then releases it momentarily to send a subsonic signal of 150 KW power to the annulus of the oil well. The echo reflected by the After the receiving device is amplified, it undergoes a series of software processes such as digital filtering and fast Fourier transform to convert it into a valid digital signal, and inputs the signal to the microprocessor through the data acquisition interface together with the signal diagram and other operating state parameters. 2, the intelligent control part consists of high-capacity microprocessor and data processing and control software, the core technology for the establishment of dynamic liquid surface echo identification model and calculation method and the preparation of the calculation software. After receiving the acoustic reflection signal, the microprocessor performs arithmetic processing to determine the liquid surface depth and sends a reference instruction according to the preset submergence. 3, the executive part of the control instructions issued by the microprocessor on the pumping unit operating parameters in real time speed control pumping strength, to achieve scheduling. Through the default multi-function interface circuit board AC variable frequency speed control, variable pole motor differential speed (256), mechanical stepless speed regulation and slip motor speed control. 4. The remote communication part borrows the GSM / GPRS network system to realize the real-time control and management of data acquisition, processing, digital wireless transmission, video monitoring and parameter adjustment of the mechanical pumping well on-site. The operation cost is lower and the work reliable. Field Tests Since 2004, four wells in the Bonan Oilfield have been tested to verify the accuracy of the test results, the reliability of the hardware and software, on-site applicability and economic feasibility. First of all, the calibration of the liquid level test was carried out. The real-time liquid level test device and the Jinggangshan echosounder were used to monitor the same liquid level at the same time. At 1840 meters, the difference between the two instruments was 30 meters, which could meet the production requirements. The following figure shows the result of the reflected wave playback. The vertical line shows the position of the echo automatically interpreted. Secondly, the stability and adaptability of the system were evaluated. Implement 5 wells. Because of the low formation permeability (10-3010-3μm2) and low production capacity, the pumping system adopts some optimization measures such as deep pumping of small pumps, long-stroke, slow-stroke, anchoring of tubing, After the system efficiency is low, the scheduling problem is not fundamentally eliminated. The above parameters are in the pump shortly after pump condition is better, no wear data. The average moving surface is 1780 m and the average pump hanging depth is 1800 m. The depth of the pump hanging is basically the same as that of the dynamic liquid surface, which reflects the inherent laws of the low permeability oilfield. The potential for increasing fluid pressure and amplifying the production pressure to increase the liquid production is very limited [ 2], the average pump efficiency of 34.4%. Deep well pump insufficient supply of the phenomenon is very common. According to the law of site statistics, the newly cast deep well pump generally experienced serious leakage after about 6 months under the condition of insufficient liquid supply. The oil production volume decreased, the dynamic liquid level rose, and the pump efficiency and system efficiency continued to drop to a low level becoming steady. After the intelligent oil recovery technology is applied, the control program is initialized and the submergence of 100-200 meters is set according to the minimum submergence requirement of the rod pump. The well is tested for hydraulics by a closed-loop control system. The actual subsidence (La) is compared with the set subsidence (Lp). When La is less than Lp, the momentum is reduced and vice versa. The five wells involved in the test were adjusted down to about 1-2 impulse / min, and the moving liquid level was stable at around 1700m. The figure below shows the comparison of the Kinetic chart before and after 5-11-1 test. Test before the test chart after the test Kungfu: 1854m Moving liquid: 1705m Maximum load: 79.7KN Maximum load: 56.0KN Minimum load: 30.9KN Minimum load: 35.6KN Third, the effect of analysis Through the four wells Test, see the following results. 1, through intelligent control, deep well pump submergence remained at 200 ~ 300 meters within the premise of a stable total output, to achieve the supply and adjustment. 2, the red down, pump fullness increased, pump efficiency increased by 17.3%. Impact load drop, balance increased, machine, rod, pump operating conditions have been improved. 3, Single well daily electricity 43.7KWh, the average system efficiency increased by 10.8%. Analysis of the effect of intelligent closed-loop oil production control system IV. CONCLUSIONS 1. The intelligent design of rod pump has the correct design concept, advanced key technologies, reliable equipment performance and good investment returns. 2, The technology integrates the new achievements of multi-disciplines such as automation, information and intelligence, and is a new development of optimization technology of sucker rod production system. In the production process, through continuous adjustment of the system online, the submerged depth of the sucker rod pump remains within a reasonable range and maximizes the potential of the oil layer so that the contradiction between supply and demand can be better solved. 3. The established model and calculation method of dynamic liquid surface echo and the supporting calculation and control software have proved that the method is correct and of high accuracy. 4, the experiment that the technology for low permeability oilfield to improve system efficiency, extend pump inspection cycle, save energy will play a very good role.