Abstract: In order to promote the transition from periodic planned preventive maintenance of harmonious-type locomotives to digital precision maintenance, and prevent major accidents and failures, this paper, based on big data and cloud platform technology, adopted Open System Architecture of Condition-Based Maintenance, OSA-CBM), designed the fault Prediction and Health Management(PHM) system of harmonious-type locomotives, including the overall architecture, technical architecture, functional architecture. It aims to master the performance evolution rules of locomotives and important parts, ensure the safety of locomotives in transit, implement accurate maintenance, and reduce the operation and maintenance costs of the whole life cycle.