Research On Unmanned Helicopter Sensor Fault Diagnosis And Fault-Tolerant Control Method

At present, the development of UAV is like a raging fire. unmanned helicopter, due to the advantages of vertical take-off and landing, slow flight, and hovering, gains much more application and more rapid development. Each main research-institution, domestic and abroad, is competing now in research and development of new types of unmanned helicopter, for civil or military use.But because of the poor working environment for the unmanned helicopter platform, such as strong interference, strong vibration. The flight controller has higher requirements of the sensor. Once a fault of the sensor occurs, it will cause the error control of unmanned helicopter or even a crash due to out of control. So, aiming the unmanned helicopter flight control system, sensor fault diagnosis and fault tolerant control occur. The flight control system of unmanned helicopter-aimed sensor fault diagnosis and fault-tolerant control has a lot of kinds of methods, at present, the most common way is the sensor redundancy design, once the sensor fault occurs, a redundant sensor will be replaced instead of the sensor with fault, and fault tolerant control is realized in this way. Due to the limit of the cost of unmanned helicopter and the limit of its load, the redundant fault diagnosis and fault tolerant control system increase the cost, also reduce the ability of the execution of the task. So this paper presents a new method of fault diagnosis and fault tolerant control method which based on a new analytical model. The method can be achieved by only one set of sensor, greatly reduces the cost of unmanned helicopter. The main research process are as follows:First, establish a mathematical model for a certain type of unmanned helicopter, it makes a preparation for the simulated verification for the fault diagnosis and fault tolerant control method. And analyzes the working principle of the IMU module of sensor for unmanned helicopter, analysis and the establishment of common fault model of IMU sensor. Further propose sensor fault diagnosis and fault tolerant control method for unmanned helicopter control system: using the established unmanned helicopter model, design the full dimensional subjective sensors and the reduce dimensional sensors of each output, to realize the decoupling of the unmanned helicopter model.Calculate the main residual of full dimensional subjective sensors and unmanned helicopter model’s output, and calculate the inferior residual of the decoupling output of the reduced-dimensional observer and the output of unmanned helicopter model. first judging the main residual with the application of the sequential probability ratio criterion, determine whether the system failure. Once identified failure occurs, we compare the inferior residuals, we will find the output of the inferior residual of the reduced-dimensional observer changes obviously only when this observer is corresponding to the faulted sensor, other inferior residuals are almost zero. This will determine the fault sensor which must be immediately isolated. using other sensor weighted output Numerical(reconstructed signal) to replace the fault sensor’s output Numerical. Then the reconstructed signal will be feedback to the unmanned helicopter flight controller. Unmanned helicopter flight control algorithm using sliding model structure-variation control algorithm for the model, So for each kind of sensor default model, it will be switched to the corresponding sliding surface control, achieve better fault tolerant control. Finally, using Matlab GUI environment to develop a simple simulation system, to complete the simulation of the algorithm. During the simulation, respectively assume that a failure has occurred for every attitude angle sensor, in order to verify the validity and practicability of the method. The simulation results show that the sensor fault diagnosis and fault tolerant control method are feasible, and relative higher time-effective.