The invention relates to a thermal power generating unit load distribution method based on a neural network and an intelligent optimization algorithm, and the method comprises the steps of obtaining the historical operation data of each unit from a distributed control system of a power plant, and obtaining the historical power supply coal consumption data from a plant-level real-time monitoring information system of the power plant; arranging the historical operation data and the historical power supply coal consumption data into a matrix form to serve as input data of a BP neural network input layer, wherein a hidden layer of the BP neural network adopts a Sigmoid function, an output layer adopts a linear function, and a weight training algorithm adopts an L-M optimization algorithm; adopting a heuristic intelligent optimization algorithm as an optimization main program, adopting the BP neural network output trained in the step 2 in the optimization main program as a fitness functionof an intelligent optimization algorithm, and screening the data operated in the optimization main program until the optimal load bearing capacity or iteration frequency of each unit of the power plant reaches a limited maximum value, so that the load optimization distribution of the thermal power generating unit is realized.