[0017] See figure 1 The device for estimating the dynamic weight of a toll station vehicle of the present invention includes a vehicle speed measurement module and an impact force measurement module. The vehicle speed measurement module is composed of a laser transmitter group and a photoelectric receiver group. The laser transmitter group and the photoelectric receiver group are respectively installed on the left and right sides of the toll station road. The laser transmitter group is composed of a first laser transmitter 2 and a second laser transmitter 4. The distance between the first laser transmitter 2 and the second laser transmitter 4 along the length of the road is L, and the photoelectric receiver group is received by the first photoelectric receiver The distance between the first and second photoelectric receivers 3 and 5 is also L along the length of the road. The first laser transmitter 2 is facing the first photoelectric receiver 3, and the emitted laser light is received by the first photoelectric receiver 3; the second laser transmitter 4 is facing the second photoelectric receiver 5, and the laser light emitted by it is The second photoelectric receiver 5 receives. The impact force measuring device 6 is placed on the road at a certain distance in front of the second laser transmitter 4 and the second photoelectric receiver 5. The outer surface of the impact force measuring device 6 is made of rubber with good elasticity and processed to be similar to The shape of the road speed bump. When the vehicle under test 1 enters the speed measurement area, the vehicle under test 1 first blocks the laser light emitted by the first laser transmitter 2 so that the first photoelectric receiver 3 cannot receive the signal; when the vehicle under test 1 is about to leave the speed measurement area In the area, the vehicle under test 1 shields the laser light emitted by the second laser transmitter 4, so that the second photoelectric receiver 5 cannot receive the signal; when the vehicle under test 1 leaves the speed measurement area, keep the original speed The impact force measuring device 6 in the shape of the speed bump in the front.
[0018] See figure 2 , Along the width of the road, a force plate is placed on the left and right sides of the inner rear of the impact force measuring device 6, which are the left force plate 17 and the right force plate 18. In order to improve the accuracy of obtaining the impact force signal, The left force plate 17 and the right force plate 18 are evenly divided into 3 areas from left to right, so as to ensure that each area can obtain the impact force F of the vehicle when the vehicle passes the impact force measuring device 6. A pressure sensor is installed below the middle of each area, from left to right: the left first pressure sensor 17a, the left second pressure sensor 17b, the left third pressure sensor 17c and the right force under the left force plate 17 The right first pressure sensor 18a, the right second pressure sensor 18b, and the right third pressure sensor 18c are under the plate 18.
[0019] See image 3 , The cross section of the impact force measuring device 6 is a trapezoid protruding upwards of the road, with For slope, the impact force measuring device 6 is provided with a right force plate 18, elastic sensitive elements 8 and 4 resistance strain gauges at the back of the interior. The right force plate 18 is placed obliquely, with the upper surface facing the rear and the front and back of the right force plate 18 The parts are respectively fixed and suspended on the road surface by the front side fixing bracket 16 and the rear side fixing bracket 11 to ensure that the right load bearing plate 18 does not move during the impact. In particular, in order to reduce the loss in the transmission process of the impact force F when the vehicle passes, it is necessary to make the right force plate 18 fit the inner surface of the impact measuring device 6 as much as possible, and the right force plate 18 is inclined at an angle θ It is designed according to the slope of the impact measuring device 6. Is the best. An elastic sensitive element 8 is closely attached to the lower surface of the middle of the right stress plate 18. The elastic sensitive element 8 is a solid rectangular block made of a constant elastic alloy material. The lower part of the elastic sensitive element 8 is fixed to the ground by a fixing bracket 13. The first resistance strain gauge 9 and the second resistance strain gauge 10 are pasted on both sides of the upper part of the elastic sensitive element 8. At the same time, the third resistance strain gauge 12 and the third resistance strain gauge 12 and The fourth resistance strain gauge 14. Four resistance strain gauges are connected to the measuring circuit 15 to form a full bridge circuit. An elastic sensitive element 8, four resistance strain gauges and a measuring circuit 15 together form a pressure sensor. An MCU control circuit box 19 is installed inside the impact force measuring device 6, and the MCU control circuit box 19 can be placed on the road. The MCU control circuit box 19 is provided with an MCU control system.
[0020] image 3 Only the installation structure and requirements of the right force plate 18 and the structure of one of the pressure sensors are shown, figure 2 The installation structure and requirements of the center left stress plate 17 image 3 The right bearing plate 18 in the middle is the same, figure 2 The structure of other pressure sensors in image 3 The structure of the pressure sensor is the same. Each pressure sensor is composed of an elastic sensitive element 8, 4 resistance strain gauges, and a measuring circuit 15. There are three pressure sensors under the left force plate 17, and these three pressure sensors form the left pressure sensor group. There are three pressure sensors under the right force plate 18, and these three pressure sensors form the right pressure sensor group. The output lines of the left and right pressure sensor groups are connected to the MCU control circuit box 19 so that the MCU control system can process the output signal of each pressure sensor.
[0021] See Figure 4 , The MCU control circuit box 19 has an integrated circuit board (not shown in the figure) composed of MCU control system, A/D conversion circuit, amplifier circuit, communication module and other related circuits, with the MCU control system as the core. The output lines of the left and right pressure sensor groups are connected to the amplifier circuit, and then connected to the input end of the MCU control system through the A/D conversion circuit. The MCU control system is connected to the external host computer through the communication module, and the output of the MCU control system is connected to the laser transmitter. The outputs of the first laser transmitter 2 and the second laser transmitter 4 in the photoelectric receiver group, and the first photoelectric receiver 3 and the second photoelectric receiver 5 in the photoelectric receiver group are connected to the input of the MCU control system. There is also a power supply module inside the MCU control circuit box 19 to supply power to the MCU control system and the left and right pressure sensor groups.
[0022] See Figure 5 When the vehicle dynamic weighing estimating device of the present invention is working, the vehicle to be tested 1 enters the speed measurement area, blocking the laser light emitted by the first laser transmitter 2, so that the output signal of the first photoelectric receiver 3 is discontinuous, and the MCU control system recognizes The signal of the first photoelectric receiver 3 is discontinuous and the timer is turned on. When the vehicle 1 under test is about to leave the speed measurement area, the laser light emitted by the second laser transmitter 4 is blocked, so that the signal output by the second photoreceiver 5 is discontinuous, and the MCU control system recognizes the signal of the second photoreceiver 5 If is not continuous, the timer is turned off, and the MCU control system thus obtains the travel time of the vehicle 1 under test as T, that is, obtains the time interval T between two occlusions. Since the distance between the first and second laser transmitters 2, 4 along the length of the road is L, the MCU control system obtains that the speed of the vehicle 1 under test before passing the impact force measuring device 6 is V=L/T, and the The speed V is transmitted to the upper computer through the communication module. When the vehicle under test 1 travels along a predetermined route and hits the impact force measuring device 6 at a certain speed V, the left and right bearing plates 17, 18 are impacted by the left and right wheels respectively, so that the bearing plates 17, 18 are installed on the left and right sides. The elastic sensitive element 8 below 18 is deformed, which causes the four resistance strain gauges attached to the elastic sensitive element 8 to deform, which causes the resistance value of the resistance strain gauge to change. The change in the resistance of the four resistance strain gauges makes the measurement circuit The full bridge circuit composed of 15 loses balance, and thus outputs a voltage signal that has a certain relationship with the impact force F. The voltage signal is amplified by the amplifier circuit and the voltage signal is converted into a digital signal by the A/D conversion circuit and then input to the MCU control system. The MCU control system transmits the collected speed V and impact force F to the upper computer through the communication module, and processes the collected data V and F through the upper computer software using artificial neural network algorithms to measure the vehicle quality M.
[0023] When the vehicle under test 1 hits the force plates 17 and 18 on the left and right sides, the MCU control system collects the signals of the left and right pressure sensors, and performs corresponding processing to obtain the impact force F. The impact force F is the average value measured by the left and right pressure sensor groups. If the impact forces measured by the three pressure sensors 17a, 17b, and 17c on the left are , Then the impact force of the vehicle under test 1 hitting the left side 17 Take the average, that is , The impact forces measured by the three pressure sensors 18a, 18b, and 18c on the right are respectively , The impact force of the vehicle under test 1 hitting the right side force plate 18 Take the average, that is , Then the impact force generated when the vehicle 1 to be tested hits the impact force measuring device 6 is. In addition, the design length d of the impact force measuring device 6 should be greater than the width of the body of the vehicle 1 to be tested (see figure 2 ), and in order to be able to take into account the width of the wheels of different vehicles or the width of the double wheels of trucks as much as possible, and to consider the possibility that the vehicle does not follow the established route or the wheels deviate during impact, the force plates on the left and right sides The width m of 17, 18 should be greater than twice the width of the wheel of the vehicle under test 1 to ensure that the pressure sensor can obtain a more accurate impact signal.
[0024] In the process of vehicle dynamic weighing, the impact force F generated by the impact force measuring device 6 of the vehicle under test 1 is affected by many factors, such as the slope of the impact force measuring device 6 and the tilt angle of the force plate θ, vehicle speed and loss during the transmission of impact force F, vibration of the vehicle itself, etc. When estimating the mass of the vehicle, the present invention selects parameters that have a relatively large impact on the measurement accuracy, that is, the speed V and the impact force F are important parameters for estimation.
[0025] See Image 6 When the MCU control system transmits the collected speed V and impact force F to the upper computer through the communication module, the upper computer uses the BP artificial neural network algorithm to process the collected degree V and impact force F. The specific processing method is:
[0026] According to the kinetic energy theorem formula And Newton’s laws of motion mechanics, where, m It is the mass of the object and v is the speed of the object. At a certain speed, the greater the mass of the object, the greater the inertia and the greater the kinetic energy of the object. Therefore, during the collision, the moving object receives more impact. The kinetic energy of the moving object There is a certain relationship with the impact force generated during the collision. It can be seen that there is a corresponding functional relationship among the mass M, speed V and impact force F of the moving vehicle 1 under test, namely.
[0027] The invention uses BP artificial neural network technology to fit the functional relationship among the mass M, the speed V and the impact force F of the moving vehicle. The speed V of the vehicle under test 1 and the impact force received by the vehicle when it hits the impact force measuring device 6 are used as the input of the BP artificial neural network model, and the mass M of the vehicle under test 1 is used as the output of the neural network model, which can construct a binary function. To fit the functional relationship of the three, it is necessary to obtain sufficient sample data. The present invention first obtains sufficient experiments through the experimental method that vehicles of different masses collide with the impact force measuring device 6 at different speeds and obtain the corresponding impact force F. Sample data, and then select part of the experimental sample data as generalized sample data, and other data as training sample data, and set the error target value and learning factor for network training to obtain the weights and thresholds after training. According to the obtained parameters, judge whether the generalization result meets the system's error requirements. If it does, save the corresponding network model. If it does not meet the system error requirements, modify the learning factor and train the network again until it meets the system error requirements. Until the request of the error. Through the trained BP artificial neural network model, the collected velocity V of the vehicle under test 1 and the impact force F received by the vehicle when it hits the impact force measuring device 6 are used as the input of the BP artificial neural network model, and the mass of the vehicle under test 1 is M As the output of the BP artificial neural network model, a more accurate vehicle mass M can be estimated.
