中文摘要：

基于引力位理论推导出长方体外任意一点的重力异常公式,将模拟的研究区域划分成许多具有固定长、宽、高且密度均匀的长方体,并给定了每个长方体的密度异常,利用重力的可叠加性,计算了重力异常分布。在此基础上,采用粒子群算法进行密度反演试验,同时将粒子群算法与其它算法进行了比较,结果表明：在适应度值为0.99的条件下,粒子群算法比遗传算法耗时减少了93.7%。

英文摘要：

This paper firstly derives the gravity anomaly formula of any point outside the Cuboid according to the gravitational theory,and then divides the simulation study area into many cuboids with a fixed length,width,height and uniform density,and sets the density anomaly of each cuboid.Meanwhile,the distribution of gravity anomaly is calculated according to the superposition of gravity.Based on the above processes,the paper makes density inversion test by means of Particle Swarm Optimization（PSO） Algorithm,and compares the results with that obtained by Genetic Algorithm（GA）.It is proved that the time consumption of PSO Algorithm is decreased by 93.7% when the fitness value is 0.99.