In order to solve the problem of optimal sensor placement in the field of wharf structure health monitoring, an optimal sensor placement algorithm based on improved multi-objective particle swarm optimization was proposed. In order to solve the problems of low optimization efficiency and single optimization objective of traditional methods, which was difficult to meet the complex health monitoring requirements such as modal identification and damage identification at the same time, a multi-objective optimization function based on damage sensitivity and redundancy, damage identification ill-posedness and modal linear independence was constructed, and the multi-objective particle swarm optimization algorithm (IMOPSO) was improved to obtain Pareto solution set, and the optimal sensor placement scheme was determined using TOPSIS entropy method. The test on a high-piled wharf shows that compared with effective independence method and effective independence-modal kinetic energy method, the distribution of measuring points using IMOPSO is more uniform, and the condition number of sensitivity matrix, the maximum non-diagonal element of MAC and the index of damage redundancy are optimized more than 45%, 90% and 5% respectively; The accuracy of damage location and degree identification under various working conditions is improved more than 5% and 7% respectively.