The　article,　based　on　satisfying　robustness　of　the　system　and　put　forward　the　objective　function　of　time-domain　performance　and　dynamic　characteristics,　introduced　genetic　operators　into　Particle　Swarm　Optimization.　The　algorithm　improve　the　diversity　of　particles　by　selection　and　hybridization　operations　and　strengthen　the　excellent　characteristics　of　particles　in　the　swarm　by　introducing　crossover　and　mutation　genes,　which　can　avoid　bog　down　into　local　optima　and　premature　convergence　and　enhance　searching　efficiency.　The　simulation　results　indicate　that　when　the　algorithm　is　applied　to　the　optimization　of　PD　controller　parameters　of　servo　system　of　grinding　wheel　rack　of　MKS8332A　CNC　camshaft　grinder,　its　performance　is　better　than　the　single　Genetic　Algorithms　or　Particle　Swarm　Optimization,　and　it　can　also　satisfy　the　demand　of　rapidity,　stability　and　robustness.