Artificial intelligence based programmable controller software designing

摘要

To automatically create software for programmable controllers for industrial plant control, a system has been developed that extracts information necessary for programming from system design specifications (such as an operation specification and actuator list) and creates programs by part retrieval. The inference engine uses a frame description language to execute a production system. Part retrieval rules are provided hierarchically and executed by the production system. When the system retrieves a software part by part retrieval rules, it retrieves the signal defined in the part from the input output signals, customizes the part, and creates an intermediate program, which is converted to a program in the target programmable controller's language. The system was applied to the creation of control programs for gradually cooling equipment in a wire and rod mill plant.