Abstract: In this paper, the start-stop control algorithm of emulsion pump based on BP neural network is studied. The algorithm selects the key parameters of liquid level, liquid temperature, oil level, oil temperature, system pressure, outlet pressure of booster pump, and cumulative running time of pump to establish a BP neural network model, to reasonably control the start and stop of the emulsion pump, MATLAB software is used for training, and the actual value is compared with the predicted value, and the results are consistent. Compared with the traditional experience of relying on operators or simply using the cumulative running time of the pump to control the start and stop of the pump, it reduces the probability of employees making mistakes, provides a new means for the reasonable start and stop of emulsion pump, and has a positive significance for improving production efficiency and saving energy consumption in coal mines.