Method for training coal rock interface recognition model, method for cutting control of shearer, and intelligent control system

A method for cutting control of a shearer, performed by an edge processor, comprising:

• receiving a target coal rock interface recognition model from a cloud server, wherein the target coal rock interface recognition model is trained by the cloud server based on a sample coal rock distribution and sample multi-modal data, and the sample multi-modal data comprises:
• sample current data of a cutting motor, sample pressure data of a lift cylinder, sample vibration data of a cutting arm, sample noise data during cutting coal rocks and sample video data for drum cutting;
• acquiring current data of a cutting motor, pressure data of a lift cylinder, vibration data of a cutting arm, noise data during cutting coal rocks and video data for drum cutting;
• acquiring a load state feature based on the vibration data and the noise data;
• acquiring a cutting feature of a coal rock interface based on the video data;
• acquiring a load feature for drum cutting based on the current data and the pressure data;
• calling the target coal rock interface recognition model, and generating a predicted coal rock distribution based on the load state feature, the cutting feature and the load feature;
• determining a target drum height and a target traction speed based on the predicted coal rock distribution;
• and
• sending the target drum height and the target traction speed to a controller of the shearer, so that a shearer is controlled to cut a coal seam.

An intelligent control system, comprising:

• a shear, configured to cut a coal seam;
• a controller, configured to control a shear to cut the coal seam;
• a current sensor, configured to acquire current data of a cutting motor;
• a pressure sensor, configured to acquire pressure data of a lift cylinder;
• a vibration sensor, configured to acquire vibration data of a cutting arm;
• a sound sensor, configured to acquire noise data during cutting coal rocks;
• a video acquisition device, configured to acquire video data for drum cutting;
• a cloud server, configured to obtain a target coal rock interface recognition model by training a coal rock interface recognition model based on a sample coal rock distribution and sample multi-modal data from an edge processor, wherein the sample multi-modal data comprises:
• sample current data of the cutting motor, sample pressure data of the lift cylinder, sample vibration data of the cutting arm, sample noise data during cutting coal rocks and sample video data for drum cutting;
• and
• an edge processor, configured to acquire a load state feature based on the vibration data and the noise data;
• acquire a cutting feature of a coal rock interface based on the video data;
• acquire a load feature for drum cutting based on the current data and the pressure data;
• receive the target coal rock interface recognition model from the cloud server, and generate a predicted coal rock distribution based on the load state feature, the cutting feature and the load feature;
• determine a target drum height and a target traction speed based on the predicted coal rock distribution;
• and send the target drum height and the target traction speed to the controller.