Scheduling system and method using a knowledge graph

A scheduling system using a knowledge graph, which is applicable to a plurality of products, with each of the plurality of products having a product part number and the scheduling system comprising:

• an input interface configured to receive an initial scheduling solution, wherein a data structure of the initial scheduling solution comprises a plurality of rows, each of the plurality of rows records at least one of a process number, a machine number, a product part number, a station name, a version, a machine name, a start time and an end time;
• a production system configured to perform the initial scheduling solution, wherein the production system comprises a plurality of stations, each of the plurality of stations comprises at least one machines, each of said at least one machines has a machine number, and said at least one machine performs production operations on the plurality of products;
• a communication device communicably connecting to the plurality of stations to receive a production resource of the production system performing the initial scheduling solution;
• a processor communicably connecting to the input interface and the communication device, wherein the processor is configured to perform a plurality of instructions to generate a target scheduling solution, and the plurality of instructions comprises:
• obtaining the initial scheduling solution from the input interface;
• converting the initial scheduling solution into a plurality of triples to form the knowledge graph, wherein each of the plurality of triples includes two entities, the two entities indicate two production resources and a relationship of the production system, the relationship is associated with another production resource of the production system performing the initial scheduling solution;
• embedding the plurality of triples of the knowledge graph into a vector space to generate an embedded vector of the plurality of production resources by a knowledge graph embedding technique;
• generating a plurality of embedded vector combinations of production resources according to the plurality of embedded vectors of production resources and computing a distance of each of the plurality of embedded vector combinations;
• and
• performing a scheduling algorithm to generate a target scheduling solution according to the plurality of embedded vector combinations, and providing reference information when the scheduling algorithm generates a schedule of a station, wherein the station is associated with at least one of the plurality of machines, and the reference information comprises at least one of the plurality of embedded vector combinations associated with the station, and the distance corresponding to said at least one embedded vector combination;
• wherein the scheduling algorithm comprise:
• selecting the station from a plurality of scheduling candidate stations, wherein the station is one of the plurality of scheduling candidate stations;
• selecting a best combination from the plurality of embedded vector combinations according to a current time and the reference information;
• and
• determining whether a number of the plurality of scheduling candidate stations is smaller than a threshold, wherein:
• outputting the scheduling solution when the number of the plurality of scheduling candidate stations is smaller than the threshold;
• and
• selecting another best combination from the plurality of embedded vector combinations according to another current time and the reference information when the plurality of scheduling candidate stations is not smaller than the threshold.