Due to the different models of reducers and motors used in different working surface chain plate conveyors, the interfaces for sensor installation will also change. Therefore, determine the installation location of the reducer sensor after thorough investigation. Due to the special environment of the working surface chain plate conveyor, the sensor will inevitably be collided or damaged. In order to ensure that the sparks generated when the sensor is damaged (mainly refers to the sensor signal line and circuit being exposed and leaking outside), it will not cause the sensor where it is located. When an explosion occurs in an explosive gas environment, both the sensor power supply and the transmission signal need to meet intrinsic safety requirements. That is to say, the sensor itself should be at least an intrinsically safe sensor, and the power supply of the sensor should meet intrinsically safe requirements.

Fault diagnosis is to judge the operating status or abnormal conditions of the chain conveyor. It has two meanings. One is to predict and forecast the operating status of the conveying equipment before the chain conveyor fails; the other is to make predictions on the location, cause, type and extent of the failure after the equipment fails. judge and make maintenance decisions. Its main tasks include fault detection, identification, evaluation, estimation and decision-making. Fault diagnosis methods include two categories: fault diagnosis methods based on mathematical models and fault diagnosis methods based on artificial intelligence. The fault diagnosis method based on neural network and information fusion technology explains the basic principles of neural network and information fusion. At the same time, examples of fault diagnosis based on neural network and fault diagnosis based on evidence theory are given.

The neural network of chain plate conveyor can be divided into two categories according to the different connection methods between neurons: feedback-free forward network and mutual combination network. The feedback-free forward network consists of an input layer, an intermediate layer and an output layer. The intermediate layer can be composed of several layers, and the neurons in each layer can only receive the output of the neurons in the previous layer. There can be a connection between any two neurons in the interconnected network, and the input signal must be repeatedly transmitted back and forth between the neurons. After several changes, the chain conveyor tends to a certain stable state or enters periodic oscillation and other other state.