In order to acquire better multi-sensor detection alliance schemes, an improved plant cell swarm algorithm is proposed. The algorithm is proposed based on the natural phenomenon that the whole plant has the maximum contact area with the sunlight due to the sunward nature of the plant in the growth process, which includes five steps: initialization of plant cell population, determination of the location of the strongest light, distribution of auxin, calculation of growth rate and update of plant cell population. The simulation part compares the improved plant cell swarm algorithm (IPCCA) with the basic plant cell swarm algorithm (BPCCA), particle swarm algorithm (PSOA), bee colony algorithm (BCOA) and wolf colony algorithm (WCOA) to verify the effectiveness of the improved plant cell swarm algorithm. The simulation results show that the improved plant cell swarm algorithm has the strongest optimization ability, can effectively avoid premature and jump out of the local optimal solution, and has good global search ability in dealing with multi-sensor detection alliance problem.