The invention provides a task allocation method for an industrial distributed data collection system. According to the method, multiple factors such as resource utilization rate of collection nodes, collection efficiency, load balance, collection reliability, heterogeneous collection nodes and the like are comprehensively considered; simulation-based initial collection node number optimization model and initial task allocation optimization model are built; and according to actual configuration and usage conditions of the collection nodes, the number of the collection nodes in operation is optimized, so that the collected resource utilization rate is increased, initial allocation of tasks including 1:1 redundant tasks and non-redundant tasks under different resource configuration and resource usage conditions of the collection nodes in the industrial distributed data collection system is realized, and the demands of industrial distributed data collection on collection timeliness, reliability, resource effective utilization and the like in an industrial big data environment are met.