The energy consumption of railway transportation is huge, and the task of energy conservation in the railway industry is arduous. How to make railway trains more energy-saving is of great significance. The locomotive energy-saving operation control system based on predictive control theory predicts the speed distance curve of train energy-saving operation by establishing the locomotive energy-saving operation speed prediction model, so as to establish the train energy-saving system. The locomotive driver controls the locomotive energy-saving operation according to the prompt information. The locomotive energy-saving operation control system is simple and reliable, and has practical reference value for electric locomotive and even industrial control.
2. System design
Figure 1 shows the structural block diagram of locomotive energy-saving operation control system based on predictive control theory. The information acquisition unit is responsible for collecting the switching value, analog value and digital value used by the information real-time calculation unit, conditioning the data through software, and finally sending the useful information to the CAN bus for other units to receive; The information real-time calculation unit receives the useful data on the CAN bus and processes the data to calculate the software call, cycle the parameters such as vehicle traction, braking force and train resistance, predict the energy-saving running speed of the locomotive, and finally send these data to the CAN bus for the information display unit to receive the data and display relevant information.
2.1 speed prediction model based on predictive control
The predictive control model predicts the future output of the system according to the historical information and future input of the controlled object. Predicting the energy-saving running speed of locomotive is to establish a model describing the dynamic characteristics of the system according to the predictive control algorithm. In the establishment of locomotive running speed prediction model, the functions of model predictive control are divided into:
(1) Historical information the line profile information on which locomotive speed prediction is based includes section slope, section length, section curve radius and length, section start and end elevation, tunnel conditions, turnout signals, etc., as well as locomotive parameters collected by the system itself, many analog, switching and digital signals, etc.
(2) The future input of the controlled object is the speed of the locomotive. In a process or cycle of speed prediction, the interval is divided into three sections: initial section, middle section and end section, and the end speed of the end section is set according to the speed limit conditions, and the self-set end speed is the future input of the controlled object.
(3) The main purpose of the future output locomotive energy-saving operation control system is to predict the speed through the known historical information, real-time calculation according to the traction calculation model and combined with the locomotive optimal control strategy. The future output of the prediction system is to predict the speed curve after calculating the interval.
2.2 rolling optimization and feedback correction of speed prediction model
The main feature of predictive control is on-line optimization. Due to the difference of the driver’s control level of the locomotive, there may be a large error between the actual running speed and the predicted speed after running for a certain distance. Therefore, the error should be corrected in time to predict the forward speed running curve at the current speed again. Although the actual speed deviates greatly from the predicted speed, after re prediction, the driver can drive at the current speed. The system optimizes the speed prediction curve by rolling, so that the system can calculate the best energy-saving operation prediction speed in the current state at any position of the locomotive. Among them, the relative form of speed prediction indexes at different times is the same, but the absolute form (i.e. the time region and location region) is different. The system can provide energy-saving operation scheme at any time and at any position, and its internal is a real-time calculation process. From the whole process of control, the optimization process is dynamic.
In addition, in the process of system speed prediction, a prediction curve is output in one cycle prediction, and there is a deviation between the actual speed at different positions of the locomotive and the predicted speed. The feedback correction of the system is to feed back the actual speed deviated from the predicted speed to the speed prediction model to renew the predicted speed curve, so as to reduce the deviation between the actual speed and the predicted curve. When the coincidence degree between the actual operation curve and the prediction curve of the locomotive is greater and greater, the energy-saving effect of the locomotive is more obvious.
3. Hardware circuit design
The hardware circuit of the system is mainly composed of information acquisition unit, information real-time calculation unit and information display unit. All units are connected via CAN bus. Among them, the information acquisition unit collects the required locomotive signals according to the traction calculation model; The real-time information calculation unit receives the collected data through the CAN bus, processes it, and predicts the energy-saving operation speed curve according to the traction calculation formula in document  and combined with the stored line profile information. The above calculation is realized by software. Figure 2 shows the interface circuit between CF card for data storage and MCU P89c591. CF card has small volume, large storage capacity, easy data storage, can store a large number of line profile data, reliable operation, low power consumption, and easy to connect with MCU. The information display module communicates with the energy-saving control information real-time calculation unit through the CAN bus, discriminates and analyzes the data sent to the CAN bus by the information acquisition unit and the information real-time calculation unit on the CAN bus, and displays the important data of locomotive operation and speed prediction curve. The system uses P87C591 single chip microcomputer as the core controller, and a powerful can controller module is integrated in it. Figure 3 shows the communication circuit between P87C591 and can bus.
4. Software design of speed prediction function
In the prediction process of locomotive energy-saving operation speed, the line profile data, calculated train force data, train operation conditions and other state signals are required. Among them, the line profile information is pre stored in the program for retrieval at any time. Figure 4 shows the flow chart of locomotive energy-saving operation speed prediction program.
P87C591 is used as the core controller, and the overall design of locomotive energy-saving operation control system is carried out based on predictive control theory. The main characteristics of the predictive control are predictive model, rolling optimization and feedback correction. The predictive control is used to accurately predict the energy-saving running speed of the locomotive, so as to better guide the locomotive driver to operate the locomotive in an energy-saving way. The system has good practical value in locomotive energy-saving control. At the same time, it also has a certain reference significance for the research of relevant energy-saving systems.
Responsible editor: GT