A　new　multi-objective　predictive　control　system　based　on　the　neural　network　was　proposed　for　activated　sludge　wastewater　treatment　process.　Both　the　dissolved　oxygen　level　in　the　fifth　compartment　and　the　nitrate　level　in　the　second　anoxic　tank　were　controlled　in　the　benchmark　simulation　model　No.　1　(BSM1).　The　control　system　mainly　consists　of　two　parts:　one　part　is　the　identifier　used　to　predict　the　outputs　of　the　object;　another　part　is　the　controller　applied　to　control　the　variables　of　the　system.　The　simulation　results　demonstrate　the　proposed　control　system　owning　better　adaptability　and　greater　robustness.