Research On The Posture Intelligent Sensing System Of The Hydraulic Support And The Decision-Making Based On The BP Neural Network

In coal mine production,it is of great significance to improve the intelligent level of coal mine equipment for safe and efficient production due to the harsh working face environment and difficult mining conditions.In this context,the optical fiber posture intelligent sensing system of the support is constructed by using the two-column shield hydraulic support as the basic equipment,and the attitude decision problem of the top beam of the support is deeply studied in this paper.The main contents of the study are as follows:(1)First,analyzing the structure,classification,sensing principle and monitoring principle of temperature and stress(strain)of optical fiber sensing technology.Then,through the analysis of the type of support and the structure characteristics of the two-prop shield hydraulic support,the posture model of the support is established and solved by vector closed-loop theory.Finally,according to the above-mentioned theoretical analysis,the posture intelligent sensing system of the support are constructed.(2)The inducting beam FBG tilt sensor has been developed and designed.Numerical simulation was carried out by using Ansys workbench,and the performance and attitude of the support were tested by the method of laboratory research.The results show that in the range of measurement,the sensitivity of the sensor is better,the theoretical value is 51.78pm/°,the measured value is 51.25pm/°,and it has high stability and adaptability to the posture of the support.(3)The posture decision model of top beam based on BP neural network is constructed.The decision index system of the top beam is determined by analyzing the influencing factors of the support top beam and the monitoring measurement of the attitude sensing system.Combined with the theory of BP neural network,the specific parameters of the decision model are determined,that is,the number of neurons in the input layer and the output layer is 17 and 1 respectively,the number of neurons in the hidden layer is 4 ~15,and the learning rate is 0.5,Expected error set to0.00001.(4)The posture decision model of top beam based on BP neural network is simulated and analyzed in MATLAB platform.The Longde 101 working face was selected to install and arrange the hydraulic support attitude sensing system.The 81#support was collected for data acquisition in the range of 100,and then the data was used to learn and test the network model in MATLAB.The simulation results showthat trainlm is the best training function and the optimal number of neurons in the hidden layer is 10;The absolute error range of the model is 0~0.06°and the relative error range is 0%~2.1%.It is proved that the model has high prediction precision and strong stability.The average absolute error is 0.021°and the average relative error is0.70%,which accords with the requirements of decision-making,which shows the validity of the model.This thesis has78 figures,10 tables and 83 references.