Space Truss Structure Health Monitoring Based On Particle Swarm Algorithm Of Optimal Placement Of The Sensor

Large-span space grid structure is commonly used in the trademark and public buildings because of its complex structure form and large structure span. Grid structure may be damaged inevitably during construction process. When the structural damage has not been found in time, accidents can be happened at any time. Engineering accidents not only brings property damage and personal injury, it can also bring immeasurable losses to the country and society. Therefore, we need to take measures to do health monitoring to important large buildings and public buildings that are crowded by people. The real-time supervision of health state can reduce the accidents caused by structural damage accumulation.Based on the space grid structure, the thesis is written to study the optimal placement of the sensor in structural health monitoring. By studying the optimization layout principles and optimization calculation method, the improved particle swarm optimization algorithm is chosen as the calculation method which is more efficient and accurate. By writing the corresponding optimization algorithm program, we made an optimization research on the number and location of acceleration sensor.When optimizing the sensor number of space grid structure model, we first use the2-norm based on Fisher information and choose the target modal number of vertical vibration of grid structure. Then, structure strain ability is chosen as the reference index,we have determined the vertical vibration monitoring order of the grid structure within the scope of the target mode. Finally, we determined the number of needed sensors by using the MAC.When optimizing the sensor location of grid structure model, the author can respectively make optimization calculation on fitness function based on the modal confidence and deformation. The author makes an improved analysis of the influence coefficient of severity of the fitness function based of the modal confidence. By calculating with the written algorithm program, the author has evaluated on both optimization results. The result shows that the latter fitness function is better than the former one and at the same time, the final optimization result is proved to be accurate and reliable.The optimization of sensor is the foundation of health monitoring. It has important practical significance and also provide basis for solving complicated optimization problems. It is especially suitable for the sensor optimization of large-span grid structure with many nodes.