Extra Force Control Study Of Electro Hydraulic Load Simulator

Electro-hydraulic load simulator is an important test equipment which is used during new aircraft development, it can be used to simulate the aero dynamic load applied to the actuators during flight, the simulated load should change corresponding to the surface deflection angle which will also simulate the aero dynamic load change duo to air speed, altitude change; Based on that the electro-hydraulic load simulator can be used to verify the actuation system performance.The loading system and actuation system inside the electro-hydraulic load simulator are directly connected, duo to this structure characteristics. Extra-force will be generated from the electro-hydraulic loading system duo to active motion from test unit. Extra-force has characteristics of advancement, real-time and will continuously change along with actuation system status change(frequency, amplitude, start, direction switch); Extra-force will have a big impact to the loading accuracy, however loading accuracy is the key performance of load simulators. So how to eliminate the extra force, improve the loading accuracy is a critical task for electro-hydraulic load simulator study.After research status review, this article listed the current status of the study on extraforce reduction, and did comparison for each method; determined the research direction for the article.Firstly, mathematics models were built up for the simulator system, the cause and characteristics of the extra force was found by analyzing the mathematics models, and additional analysis was made to understand the relation between the key system factor and extra-force behavior. Then feed-forward compensation control was used combined with traditional PID control to reduce the extra force, and analysis result shows extra force was reduced by adding the feed-forward control compensation block.In order to further reduce the extra force, this article used RBF neural network based object identifier with single neuron PID controller to increase the loading accuracy of the load simulator, RBF neural network based object identifier was used to identify the test object, single neuron PID controller can self-study, self-adjust the parameters inside the network controller based on the control error until the error close enough to zero, based on the analysis result comparison for different control method, concluded that single neuron PID controller can further reduce the extra force compare with feed forward control and PID control.