Research On The Reception And Recognition Of Ground-Downhole Flow Pulse Signals In Oil And Gas Wells

Increasing the proven geological reserves of oil and gas and improving oil and gas production are key tasks in the oil and gas industry to ensure China’s energy security.Complex hydrocarbon reservoir contains rich oil-gas resource.Therefore,speeding up the exploitation of complex hydrocarbon accumulations can increase our country oil and gas output quickly.Engineering practice shows that rotary steerable drilling technology is an important technical means to exploit complex oil and gas reservoirs,and the downlink communication that transmits ground control instructions to downhole steerable tools is an important part of the closed-loop control of rotary steerable drilling system.Due to the restriction of drill string rotation in the drilling process,the drilling fluid flow pulse method is still the most widely used and the lowest cost downlink communication means at present,but it has the problems of low transmission rate and high error rate.So how to improve the accuracy of drilling fluid flow pulse communication is the key problem that must be solved in guiding drilling technology research.This thesis adopts the research method combining theoretical research and engineering experiment to carry out research from two aspects of improving the detection accuracy of signals and the accuracy of recognition signals.The main research contents and results are as follows:1^st.This thesis studies the principle of drilling fluid flow pulse downlink communication,establishes the drilling fluid flow pulse transmission signal chain,analyzes the main factors that affect the guidance drilling tool to recognize the downlink instruction effectively,and puts forward the measures to eliminate or reduce the error of the downlink signal.2^nd.The solution that the error of drilling fluid flow pulse signal is easy to cause code error,a small-size downhole voltage sensor that meets the requirements of downhole high temperature environment is developed,and the down-transmission signal detection and processing circuit is designed.The temperature characteristics of the sensor and its signal processing circuit are tested experimentally,and the temperature compensation method is proposed.The measurement error after compensation is less than 3%.The developed voltage sensor and its signal processing circuit have been successfully applied in a steering drilling tool.3^rd.The signal characteristics of drilling fluid flow pulse downlink communication of"three drops and three liters"were studied,and the logical judgment and recognition method of downlink instruction based on the determination of signal falling edge,signal rising edge and time width was proposed.The hydraulic drive experiment showed that the identification bit error rate of this method was high.4^th.As the logical judgment and recognition method of down-transmission commands has problems of complicated logic and high misjudgment probability,a neural network intelligent recognition algorithm for down-transmission signals is proposed.In view of the high cost,long cycle and small number of samples obtained by experimental method,a new idea of using mathematical model to obtain training sample data is put forward.By studying the characteristics of production,transmission and acquisition of drilling fluid flow pulse signal,a simulation model of drilling fluid flow pulse downlink communication system is established.5^th.The structure of the command down neural network with 6 hidden layers was designed.The simulation model of drilling fluid flow pulse downlink communication system was used to generate more than 50,000 sets of sample data for the training of the neural network.100 sets of hydraulic drive experiment samples and 100 sets of simulation experiment samples were used to test the trained model,among which all the simulation samples were correctly identified.The recognition accuracy rate is 100%,while there are only 2 groups of identification errors in the sample data of hydraulic drive experiment,and the recognition accuracy rate reaches 98%,achieving a good recognition effect.These research works can provide a theoretical reference for solving the problem of high error rate of drilling fluid flow pulse downlink communication.