A mobile intelligent transmission system

Abstract

The present invention relates to the technical field of mechanical equipment, and discloses a movable intelligent transmission system, including a movable trolley, a primary conveyor line, and a secondary conveyor line. The primary conveyor line is arranged on the movable trolley; the secondary conveyor line is connected to the end of the primary conveyor line, and the movable trolley is also provided with an adjustment mechanism for adjusting the height and angle of the primary conveyor line and the secondary conveyor line, and a plurality of drive motors are respectively arranged on the adjustment mechanism. The intelligent transmission system also includes a plurality of motor speed controllers, and the output ends of the motor speed controllers are respectively connected to the corresponding drive motors. Compared with the prior art, the present invention facilitates moving the transmission platform to a suitable position, and the motor speed controller effectively solves the problem that the motors of the existing equipment of the same type do not have the characteristics of nonlinearity, large hysteresis and complex changes of the motor speed change, and intelligently predicts and adjusts the motor speed, thereby greatly improving the system performance.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical equipment, and in particular to a movable intelligent transmission system. Background Art

[0002] With the advancement of my country's modernization and the improvement of economic level, the automation level of the processing industry in the domestic market has also improved. For some large and medium-sized processing enterprises, the degree of automation of processes from the transportation and processing of raw materials to the packaging of semi-finished products has been quite high, basically realizing the full-automatic or semi-automatic processing process from raw materials to finished products.

[0003] In the process of processing goods, it is necessary to transport goods to various links for processing through a conveying platform. In addition, it is often necessary to move a batch of goods from one place to another for processing. The traditional practice is to first gather the goods in one place and then transport them to another place by a transport vehicle, forklift, etc., and then manually move the goods to the conveying platform for transportation and processing. This process wastes a lot of manpower and material resources, and the efficiency is very low. There is an urgent need for a movable conveying platform to use its mobility to reduce labor costs.

[0004] During the use of the transmission platform, it is essential to use a drive motor to drive the relevant equipment to move or adjust it. However, the existing motors of the same type of equipment do not have the characteristics of nonlinearity, large lag and complex changes in motor speed to intelligently predict and adjust the motor speed, which greatly affects the system performance. Summary of the invention

[0005] Purpose of the invention: In view of the problems existing in the prior art, the present invention provides a movable intelligent transmission system, which facilitates the movement of the transmission platform to a suitable position through a movable trolley, thereby solving the problem that the traditional immovable transmission platform cannot adjust its position, and the items can be accurately transported to the destination through a two-stage conveyor line. In addition, the motor speed controller effectively solves the problem that the drive motor of the existing similar equipment does not have the nonlinearity, large lag and complex changes according to its speed change, and intelligently predicts and adjusts the drive speed, thereby greatly improving the system performance.

[0006] Technical solution: The present invention provides a movable intelligent transmission system, comprising a movable trolley, a primary conveyor line, and a secondary conveyor line, wherein the primary conveyor line is arranged on the movable trolley; the secondary conveyor line is connected to the end of the primary conveyor line, and the movable trolley is also provided with an adjustment mechanism for adjusting the height and angle of the primary conveyor line and the secondary conveyor line, and the goods are sequentially transmitted to the destination via the primary conveyor line and the secondary conveyor line, and the adjustment mechanism is respectively provided with a plurality of drive motors, and the intelligent transmission system also comprises a plurality of motor speed controllers, and the output ends of the motor speed controllers are respectively connected to the corresponding drive motors; the motor speed controller comprises an STM32 single-chip microcomputer, an L298 motor drive circuit and a speed sensor, the output of the intelligent controller in the STM32 single-chip microcomputer is used as the input of the L298 motor drive circuit, the output of the L298 motor drive circuit is used as the input of the controlled drive motor, and the speed sensor measures the actual speed value of the controlled drive motor; the intelligent controller comprises a DRNN neural network setting model, a DRNN neural network prediction model, a DRNN neural network real-time model, a TDNN neural network model, an Elma n neural network prediction model, hybrid neural network controller, NARX neural network controller and TDNN neural network controller; the driving motor speed set value is used as the input of the DRNN neural network setting model, the difference between the driving motor speed set value and the output of the Elman neural network prediction model is used as the input of the DRNN neural network prediction model, the difference between the driving motor speed set value and the speed sensor output value is used as the input of the DRNN neural network real-time model, the outputs of the DRNN neural network setting model, the DRNN neural network prediction model and the DRNN neural network real-time model are used as the corresponding inputs of the TDNN neural network model respectively, the difference between the TDNN neural network model output and the speed sensor output value is used as the input of the hybrid neural network controller, the sum of the hybrid neural network controller output and the NARX neural network controller output is used as the input of the NARX neural network controller and the TDNN neural network controller respectively, the output of the TDNN neural network controller is used as the corresponding input of the Elman neural network prediction model and the input of the L298 motor drive circuit respectively, and the speed sensor output value is used as the corresponding input of the Elman neural network prediction model.

[0007] Furthermore, the hybrid neural network controller is composed of a tapped delay line TDL (Tapped Delay Line), an Adaline neural network model, a DRNN neural network model and a dynamic recursive wavelet neural network model. The input of the hybrid neural network controller is used as the input of the tapped delay line TDL, the output of the tapped delay line TDL includes the input signal at the current moment and the input signal delayed by 1 to multiple time steps, the output of the tapped delay line TDL is used as the input of the Adaline neural network model and the DRNN neural network model respectively, the output of the Adaline neural network model and the DRNN neural network model is used as the corresponding input of the dynamic recursive wavelet neural network model, and the output of the dynamic recursive wavelet neural network model is used as the output of the hybrid neural network controller.

[0008] Furthermore, the adjustment mechanism includes a first-level conveyor line front end adjustment mechanism and a first-level conveyor line end adjustment mechanism;

[0009] The front end adjustment mechanism of the first-level conveyor line includes a screw adjustment component A, a disc, a pair of push rods and a conveyor line bracket. The disc is threadedly connected to one end of the screw adjustment component A. The pair of push rods are symmetrically fixed on the disc. The lower surface of the first-level conveyor line is rotatably connected to the conveyor line bracket, and the top end of the push rod is connected to the conveyor line bracket.

[0010] The first-stage conveyor line end adjustment mechanism includes a screw adjustment assembly B and bosses rotatably connected to both sides of the first-stage conveyor line end, and the screw in the screw adjustment assembly B is threadedly connected to the bosses;

[0011] Furthermore, the screw adjustment assembly A and the screw adjustment assembly B have the same structure, and both include a stepper motor and a screw, and the stepper motor is connected to its corresponding motor speed controller.

[0012] Furthermore, an anti-jumping and height-limiting mechanism is also provided at the end of the first-level conveyor line, and the anti-jumping and height-limiting mechanism includes an anti-jumping and height-limiting frame and a pair of anti-jumping and height-limiting brackets; the bottom end of the anti-jumping and height-limiting bracket is fixed on the mobile trolley, and the upper end thereof is connected to the end of the first-level conveyor line, and an anti-jumping and height-limiting frame height adjustment component is also provided on the pair of anti-jumping and height-limiting brackets.

[0013] Furthermore, the height adjustment assembly of the anti-jump height limit frame includes the following structure: grooves are respectively opened on both sides of the outer wall of the anti-jump height limit bracket along its vertical direction, and racks are slidably connected in a pair of grooves, with the tooth mouth direction of the racks facing outwards; a first driven gear is meshed on the racks, and the pair of first driven gears are connected by a first rotating shaft, and a second driven gear is coaxially arranged on the first rotating shaft, the second driven gear is meshed with the driving gear, the output shaft of the first stepper motor is connected to the driving gear, and the first stepper motor is connected to its corresponding motor speed controller.

[0014] Furthermore, the mobile vehicle includes a mobile platform, a pair of steering wheels, a pair of driving wheels, a driving mechanism and a steering mechanism. The pair of steering wheels and the pair of driving wheels are connected to the mobile platform through shafts, the driving mechanism is drivingly connected to the pair of driving wheels, and the steering mechanism is connected to the pair of steering wheels; ranging and obstacle avoidance ultrasonic sensors are also provided below the front and rear ends of the mobile platform of the mobile vehicle.

[0015] Furthermore, dust covers are provided at the front and rear ends of the mobile trolley, and the dust covers include a winding rod rotatably connected to the lower surface of the mobile platform, an isolation cover wound on the winding rod, a fifth stepper motor is provided on the winding rod, and a receiving groove directly opposite to the isolation cover is provided directly below the winding rod, a pressure sensor is provided at the bottom end of the receiving groove, and an infrared sensor is provided on the side wall of the receiving groove, which are all connected to the controller, and the output end of the controller is connected to the fifth stepper motor; when the isolation cover is lowered, the ranging obstacle avoidance ultrasonic sensor is located inside the isolation cover.

[0016] Beneficial effects:

[0017] 1. The present invention sets the entire transmission platform on a mobile trolley, which can be moved to any position for use, and transported to different positions through two-level conveyor lines for cargo transportation. The DRNN neural network setting model, DRNN neural network prediction model, DRNN neural network real-time model and TDNN neural network model optimize the motor speed setting value from the perspectives of motor given, motor speed prediction and actual value of motor speed, improve the response speed, stability and accuracy of the controlled motor speed, and improve the robustness of the motor speed control system.

[0018] 2. The NARX neural network controller of the present invention is a dynamic recursive neural network with an output feedback connection. In terms of topological connection relationship, it is equivalent to the BP neural network with input delay plus the output-to-input delay feedback connection. Its structure consists of an input layer, a delay layer, a hidden layer and an output layer, wherein the input layer nodes are used for the input of the NARX neural network controller, the delay layer nodes are used for the time delay of the input signal and the output feedback signal, the hidden layer nodes use the activation function to perform nonlinear operations on the delayed signal, and the output layer nodes are used to linearly weight the hidden layer output to obtain the final predicted output of the input signal.

[0019] 3. Use hybrid neural network controller, NARX neural network controller and TDNN neural network controller to perform dynamic and static control of motor speed and repeated joint control to improve the response speed, accuracy and robustness of motor speed control.

[0020] 4. An anti-jump height limit mechanism is set between the primary conveyor line and the secondary conveyor line. The anti-jump height limit mechanism can adjust the height. The height of the anti-jump height limit frame is adjusted according to the actual needs of the site. On the one hand, it is used to limit the height of the goods, and on the other hand, it is used to prevent the goods from jumping, thereby preventing the goods from falling.

[0021] 5. The mobile vehicle of the present invention controls steering through the front wheels and drives forward through the rear wheels. A ranging and obstacle avoidance ultrasonic sensor is used to realize the positioning and obstacle avoidance of the mobile vehicle itself, and to make reasonable position adjustments to the mobile vehicle, so that the position of the transmission platform is more reasonable.

[0022] 6. The height and angle can be adjusted through the adjustment mechanism on the primary conveyor line. The push rod and the screw rod can be used to make high-precision lifting and lowering adjustments and tilt angle adjustments. The height and angle can be adjusted according to the actual needs of the site. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a schematic diagram of the structure of the motor speed controller of the present invention;

[0024] Figure 2 It is a schematic diagram of the overall structure of the present invention;

[0025] Figure 3 This is a schematic diagram of the structure of the anti-jump height-limiting mechanism of the present invention;

[0026] Figure 4 It is an enlarged schematic diagram of the driving structure of the anti-jump height-limiting mechanism of the present invention;

[0027] Figure 5 This is an overall schematic diagram of the front end adjustment mechanism of the first-level conveyor line of the present invention;

[0028] Figure 6It is a schematic diagram of the details of the front end adjustment mechanism of the first-level conveyor line of the present invention;

[0029] Figure 7 This is a schematic diagram of the structure of the mobile trolley of the present invention;

[0030] Figure 8 This is an enlarged schematic diagram of one end of the steering wheel of the mobile trolley of the present invention;

[0031] Fig. 9 This is a schematic diagram of one end of the driving wheel of the mobile trolley of the present invention;

[0032] Fig.10 This is a schematic diagram of the structure of the isolation cover above the dust cover of the present invention;

[0033] Fig.11 It is a schematic diagram of the receiving groove structure below the dust cover of the present invention.

[0034] Among them, 1-mobile trolley, 101-mobile platform, 102-steering wheel, 103-driving wheel, 104-range obstacle avoidance ultrasonic sensor, 105-fourth stepper motor, 106-second driving gear, 107-third driven gear, 108-second rotating shaft, 109-servo, 110-winding rod, 111-isolation cover, 112-fifth stepper motor, 113-receiving slot, 114-pressure sensor, 115-infrared sensor, 2-first-level conveyor line, 201-anti-jump height limit frame , 202-anti-jump height limit bracket, 203-groove, 204-rack, 205-first driven gear, 206-first rotating shaft, 207-second driven gear, 208-first driving gear, 209-first stepper motor, 210-first screw rod, 211-support column, 212-boss, 213-second stepper motor, 214-second screw rod, 215-third stepper motor, 216-disc, 217-push rod, 218-conveyor line bracket, 219-slider, 3-secondary conveyor line. DETAILED DESCRIPTION

[0035] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0036] The present invention discloses a mobile intelligent transmission system. Figure 1 To Attachment Fig.11 , mainly including a mobile trolley 1, a primary conveyor line 2, and a secondary conveyor line 3. The primary conveyor line 2 is arranged on the mobile trolley 1; the secondary conveyor line 3 is connected to the end of the primary conveyor line 2, and the goods are transported to the destination via the primary conveyor line 2 and the secondary conveyor line 3 in sequence. Figure 1The primary conveyor line 2 and the secondary conveyor line 3 are both roller transmissions, and rollers are connected to them in turn. Each roller realizes transmission through a transmission mechanism. The goods move forward under the action of the rollers of the primary conveyor line 2 and the secondary conveyor line 3, and gradually move to other equipment. The transmission mechanism can be a chain that drives the gear shaft to rotate to realize the rolling of the roller. The mobile trolley 1 is also provided with an adjustment mechanism for adjusting the height and angle of the primary conveyor line 2 and the secondary conveyor line 3. Several drive motors are respectively arranged on the adjustment mechanism. The intelligent transmission system also includes several motor speed controllers, and the output ends of the motor speed controllers are respectively connected to the corresponding drive motors.

[0037] An anti-jump height limit mechanism is also provided at the end of the primary conveyor line 2, and the anti-jump height limit mechanism includes an anti-jump height limit frame 201 and a pair of anti-jump height limit brackets 202. The bottom end of the anti-jump height limit bracket 202 is fixed to the mobile trolley 1, and the upper end thereof is connected to the end of the primary conveyor line 2. An anti-jump height limit frame height adjustment component is provided on the anti-jump height limit bracket 202, and the two ends of the anti-jump height limit frame 201 move up and down relative to the pair of anti-jump height limit brackets 202 under the action of the anti-jump height limit frame height adjustment component.

[0038] See attached Figure 3 With attached Figure 4 The anti-jump height limit frame 201 is an arch structure, and a pair of anti-jump height limit brackets 202 are vertically fixed on the mobile trolley 1. The height adjustment assembly of the anti-jump height limit frame includes the following structure: grooves 203 are respectively provided on both sides of the outer wall of the anti-jump height limit bracket 202 along its vertical direction, and racks 204 are respectively slidably connected in the pair of grooves 203, and the tooth opening direction of the rack 204 faces outward. The racks 204 are meshed with first driven gears 205, and the pair of first driven gears 205 are connected by a first rotating shaft 206. The first rotating shaft 206 is coaxially provided with a second driven gear 207, and the second driven gear 207 is meshed with a first driving gear 208, and the output shaft of the first stepper motor 209 is connected to the first driving gear 208. In order to limit the first rotating shaft 206, a slider 219 is provided on the side wall of the anti-jump height limiting bracket 202. The first rotating shaft 206 passes through the slider 219, and the slider 219 does not affect the rotation of the first rotating shaft 206. The slider 219 can play a limiting role and limit the freedom of the first rotating shaft 206. The first stepper motor 209 rotates to drive the first driving gear 208 to rotate, and the first driving gear 208 rotates to drive the second driven gear 207 to rotate, so as to realize the rotation of the first rotating shaft 206 and drive the rack 204 to move up and down.

[0039] A support column 211 is also slidably arranged inside the anti-jumping height limit bracket 202, and the top of the support column 211 is connected to the bottom of the anti-jumping height limit frame 201. When the rack 204 drives the anti-jumping height limit frame 201 to move upward, the support column 211 slides inside the anti-jumping height limit bracket 202, thereby limiting the entire anti-jumping height limit frame 201.

[0040] An adjusting mechanism is also provided on the primary conveyor line, and the adjusting mechanism comprises an adjusting mechanism at the front end of the primary conveyor line and an adjusting mechanism at the end of the primary conveyor line.

[0041] The front end adjustment mechanism of the primary conveyor line includes a screw adjustment assembly A, a disc 216, a pair of push rods 217 and a conveyor line bracket 218. The screw adjustment assembly A includes a second screw 214 and a third stepper motor 215. The output shaft of the third stepper motor 215 is connected to the second screw 214 through a coupling. The disc 216 is threadedly connected to one end of the second screw 214. A pair of push rods 217 are symmetrically fixed on the disc 216. The lower surface of the primary conveyor line 2 is rotatably connected to the conveyor line bracket 218, and the top of the push rod 217 is connected to the conveyor line bracket 218. The second screw 214 rotates to realize the up and down movement of the disc 216, and the disc 216 drives the pair of push rods 217 to move up and down, so as to realize the angle and height adjustment of the primary conveyor line 2.

[0042] The end adjustment mechanism of the primary conveyor line includes a screw adjustment assembly B, which includes a first screw 210 and a second stepper motor 213. The first screw 210 is arranged on the anti-jump height limit bracket 202, and the first screw 210 is connected to the output shaft of the second stepper motor 213 through a coupling. The first screw 210 is threadedly connected to the bosses 212 on both sides of the end of the primary conveyor line 2, so that the end height of the primary conveying platform 2 can be adjusted while fixing the primary conveyor line 2. The boss 212 can rotate relative to the primary conveyor line 2 to ensure that the primary conveying platform 2 can be fine-tuned in angle.

[0043] The mobile car 1 used in the present invention comprises a mobile platform 101, a pair of steering wheels 102, a pair of driving wheels 103, a driving mechanism and a steering mechanism, wherein the pair of steering wheels 102 and the pair of driving wheels 103 are both rotatably connected to the mobile platform 101, the driving mechanism is drivingly connected to the pair of driving wheels 103, and the steering mechanism is connected to the pair of steering wheels 102. Distance-measuring and obstacle-avoiding ultrasonic sensors 104 are respectively provided below the front and rear ends of the mobile platform 101 of the mobile car 1, and the distance-measuring and obstacle-avoiding ultrasonic sensors 104 are connected to a controller (not shown in the figure), and the output end of the controller is connected to the driving mechanism and the steering mechanism.

[0044] In this embodiment, the driving mechanism is driven by a motor, and a fourth stepper motor 105 is arranged on the mobile platform 101. The fourth stepper motor 105 is meshed with a third driven gear 107 through a second driving gear 106 to drive the second rotating shaft 108 to rotate. Two driving wheels 103 are connected to the second rotating shaft 108 to realize the rotation of the two driving wheels 103. The steering drive is also driven by a motor, and the two steering wheels 102 are rotatably connected to the mobile platform 101, which is connected to a steering gear 109, and the steering gear 109 controls the steering of the steering wheel. The steering gear 109, the fourth stepper motor 105, and the distance measurement and obstacle avoidance ultrasonic sensor 104 are connected to a controller (not shown in the figure), and the distance measurement and obstacle avoidance ultrasonic sensor 104 returns a signal to the controller, and the controller outputs a steering signal or a driving signal to the steering gear 109 and the fourth stepper motor 105. This is a prior art and will not be repeated here.

[0045] Dust covers are also provided at the front and rear ends of the mobile trolley 1, and the dust covers include a reeling rod 110 rotatably connected to the lower surface of the mobile platform 101, an isolation cover 111 wound on the reeling rod 110, a fifth stepping motor 112 is provided on the reeling rod 110, a receiving slot 113 is provided directly below the reeling rod 110, a pressure sensor 114 is provided at the bottom of the receiving slot 113, and an infrared sensor 115 is provided on the side wall of the receiving slot 113, which are all connected to the controller, and the output end of the controller is connected to the fifth stepping motor 112. When the isolation cover 111 is pulled down, the distance measurement and obstacle avoidance ultrasonic sensor 104 is located in the isolation cover 111.

[0046] In the above structure, the first stepper motor 209, the second stepper motor 213, and the third stepper motor 215 are all connected to a corresponding motor speed controller for controlling the speed of the above drive motors.

[0047] The structure of the motor speed controller in the present invention is shown in the attached Figure 1The motor speed controller includes an STM32 single-chip microcomputer, an L298 motor drive circuit and a speed sensor. The output of the intelligent controller in the STM32 single-chip microcomputer is used as the input of the L298 motor drive circuit, and the output of the L298 motor drive circuit is used as the input of the controlled motor. In the present invention, the controlled motor is the several stepping motors mentioned above, which are also the drive motors mentioned at the beginning of this embodiment. The speed sensor measures the actual speed value of the controlled motor; the intelligent controller includes a DRNN neural network setting model, a DRNN neural network prediction model, a DRNN neural network real-time model, a TDNN neural network model, an Elman neural network prediction model, a hybrid neural network controller, a NARX neural network controller and a TDNN neural network controller; the motor speed set value is used as the input of the DRNN neural network setting model, the difference between the motor speed set value and the output of the Elman neural network prediction model is used as the input of the DRNN neural network prediction model, the difference between the motor speed set value and the speed sensor output value is used as the input of the DRNN neural network real-time model, the outputs of the DRNN neural network setting model, the DRNN neural network prediction model and the DRNN neural network real-time model are respectively used as the corresponding inputs of the TDNN neural network model, the difference between the TDNN neural network model output and the speed sensor output value is used as the input of the hybrid neural network controller, the sum of the hybrid neural network controller output and the NARX neural network controller output is respectively used as the inputs of the NARX neural network controller and the TDNN neural network controller, the output of the TDNN neural network controller is respectively used as the corresponding input of the Elman neural network prediction model and the input of the L298 motor drive circuit, and the speed sensor output value is used as the corresponding input of the Elman neural network prediction model.

[0048] The hybrid neural network controller consists of a tapped delay line TDL (Tapped Delay Line), an Adaline neural network model, a DRNN neural network model and a dynamic recursive wavelet neural network model. The input of the hybrid neural network controller is used as the input of the tapped delay line TDL. The output of the tapped delay line TDL includes the input signal at the current moment and the input signal delayed by one to multiple time steps. The output of the tapped delay line TDL is used as the input of the Adaline neural network model and the DRNN neural network model respectively. The outputs of the Adaline neural network model and the DRNN neural network model are used as the corresponding inputs of the dynamic recursive wavelet neural network model. The output of the dynamic recursive wavelet neural network model is used as the output of the hybrid neural network controller.

[0049] 1. Intelligent controller design

[0050] The intelligent controller includes a DRNN neural network setting model, a DRNN neural network prediction model, a DRNN neural network real-time model, a TDNN neural network model, an Elman neural network prediction model, a hybrid neural network controller, a NARX neural network controller and a TDNN neural network controller.

[0051] 1. DRNN neural network setting model design

[0052] The motor speed set value is used as the input of the DRNN neural network setting model. The DRNN neural network setting model is a dynamic regression neural network with feedback and the ability to adapt to time-varying characteristics. The network can more directly and vividly reflect the dynamic change performance of the motor speed set value. The DRNN neural network setting model is a 3-layer network structure, and its hidden layer is a regression layer. In the DRNN neural network setting model of this patent, let I = [I1(t), I2(t), …, I n (t)] is the network input vector, where I i (t) is the input of the i-th neuron in the input layer of the DRNN network at time t, and the output of the j-th neuron in the regression layer is X j (t), S j (t) is the sum of the inputs of the jth regression neuron, f(·) is a function of S, and O(t) is the output of the DRNN neural network setting model. The output layer output of the DRNN neural network setting model is:

[0053]

[0054] 2. Design of DRNN neural network prediction model and DRNN neural network real-time model

[0055] The difference between the motor speed set value and the output of the Elman neural network prediction model is used as the input of the DRNN neural network prediction model, the difference between the motor speed set value and the speed sensor output value is used as the input of the DRNN neural network real-time model, and the outputs of the DRNN neural network setting model, the DRNN neural network prediction model, and the DRNN neural network real-time model are used as the corresponding inputs of the TDNN neural network model respectively; the design of the DRNN neural network prediction model and the DRNN neural network real-time model refers to the design method of the DRNN neural network setting model of this patent.

[0056] 3. TDNN neural network model design

[0057] The outputs of the DRNN neural network setting model, DRNN neural network prediction model, and DRNN neural network real-time model are used as the corresponding inputs of the TDNN neural network model, and the difference between the output of the TDNN neural network model and the output value of the speed sensor is used as the input of the hybrid neural network controller; the TDNN neural network model (Time Delay Neural Networks, TDNN neural network) is a time delay neural network model. It is an adaptive linear network. Its input enters from the left side of the network, and through the action of the single-step delay line D, it becomes the input of the d+1-dimensional vector after d steps of delay. The vector is composed of the current K input signals and the d input signals before K. The neuron uses a linear activation function. The time delay neural network is a variant of the traditional artificial neural network. The time delay neural network structure consists of an input layer, an output layer, and one or more hidden layers. The neural network establishes the mapping relationship between "input-output". Different from traditional neural networks, time-delay neural networks achieve memory of previous inputs by delaying inputs at the input layer. By delaying multiple input signals at the input layer, the network can use multiple input signals of the previous d steps and the current input signal to jointly predict the output at the current time point. For a time-delay neural network with an input layer delay of d steps, R is the forward propagation operator of the time-delay neural network. The relationship between the input sequence X and the output sequence Y can be simply expressed as follows:

[0058] Y(t)=R(X(t),X(t-1),…,X(td)) (2)

[0059] 4. Elman neural network prediction model design

[0060] The output of the TDNN neural network controller is used as the corresponding input of the Elman neural network prediction model and the input of the L298 motor drive circuit, and the output value of the speed sensor is used as the corresponding input of the Elman neural network prediction model; the Elman neural network prediction model can be regarded as a forward neural network with local memory units and local feedback connections. In addition to the hidden layer, there is also a special association layer; the association layer receives feedback signals from the hidden layer, and each hidden layer node has a corresponding association layer node connection. The association layer uses the hidden layer state of the previous moment together with the network input at the current moment as the input of the hidden layer, which is equivalent to state feedback. The transfer function of the hidden layer is generally a Sigmoid function, the output layer is a linear function, and the association layer is also a linear function. In order to effectively solve the problem of the approximation accuracy of the motor speed, the role of the association layer is enhanced. Assume that the number of input layer, output layer and hidden layer of the Elman neural network prediction model are m, n and r respectively; w1, w2, w3 and w4 represent the connection weight matrices from the structural layer unit to the hidden layer, from the input layer to the hidden layer, from the hidden layer to the output layer, and from the structural layer to the output layer respectively. Then the expressions of the hidden layer, association layer and output layer of the Elman neural network prediction model are:

[0061]

[0062] c p (k) = x p (k-1) (4)

[0063]

[0064] 5. Hybrid neural network controller design

[0065] The hybrid neural network controller consists of a tapped delay line TDL (Tapped Delay Line), an Adaline neural network model, a DRNN neural network model and a dynamic recursive wavelet neural network model.

[0066] (1) Adaline neural network model design

[0067] The input of the Adaline neural network model is the output of a tapped delay line TDL (Tapped Delay Line). The output of the Adaline neural network model is used as an input of the dynamic recursive wavelet neural network model. The input signal of the hybrid neural network controller is input from the tapped delay line TDL. At the output end of the tapped delay line TDL is an R-dimensional vector P, which contains the input signal at the current moment and the input signal delayed by 1 to R-1 time steps respectively. Its transfer function is a linear function. The output of the Adaline neural network model is:

[0068] a=purelin(W·p+b)=W·p+b (6)

[0069] The Adaptive Linear Element of the Adaline neural network is one of the early neural network models. In the Adaline neural network, there is a special input, namely the ideal response output d(K). It is sent to the Adaline neural network, and then compared with the network output y(K). The difference is sent to the learning algorithm mechanism to adjust the weight vector until the optimal weight vector is obtained. y(K) and d(K) tend to be consistent. The adjustment process of the weight vector is the learning process of the network. The learning algorithm is the core part of the learning process. The weight optimization search algorithm of the Adaline neural network adopts the LMS algorithm least squares method.

[0070] (2) DRNN neural network model design

[0071] The outputs of the beat delay line TDL are used as the inputs of the Adaline neural network model and the DRNN neural network model respectively. The outputs of the Adaline neural network model and the DRNN neural network model are used as the corresponding inputs of the dynamic recursive wavelet neural network model. The design process of the DRNN neural network model refers to the design method of the DRNN neural network setting model of this patent.

[0072] (3) Dynamic recursive wavelet neural network model design

[0073] The dynamic recursive wavelet neural network model is a dynamic improvement based on the wavelet network. The dynamic wavelet recursive neural network WNN (WaveletNeuralNetworks) theoretical basis is a feedforward network proposed by combining the wavelet function as the excitation function of the neuron and the artificial neural network. The scaling, translation factor and connection weight of the wavelet in the dynamic recursive wavelet neural network are adaptively adjusted in the process of optimizing the error energy function. Assume that the input signal of the dynamic recursive wavelet neural network can be represented as the input one-dimensional vector x i (i=1,2,…,n), the output signal is represented by y k (k=1,2,…,m), the calculation formula of the dynamic recursive wavelet neural network model output is:

[0074]

[0075] In the formula, ω ij The connection weight between the input layer i node and the hidden layer j node, is the wavelet basis function, b j is the translation factor of the wavelet basis function, a j The scaling factor of the wavelet basis function, ω jkis the connection weight between the hidden layer j node and the output layer k node. The difference between the dynamic recursive wavelet neural network model of this patent and the ordinary static recursive wavelet neural network is that the dynamic recursive wavelet neural network model has two association layer nodes that store the "internal state" of the network. A self-feedback loop with a fixed gain is added to the two association layer nodes to enhance the memory performance of the time series feature information, thereby enhancing the tracking accuracy of the evolution trajectory of the motor speed control quantity to ensure better control accuracy; the first association layer node is used to store the state of the hidden layer node at the previous moment, and then pass it to the hidden layer node at the next moment; the second association layer node is used to store the state of the output layer node at the previous moment. state, and then pass it to the hidden layer nodes at the next moment; the feedback information of the neurons in the hidden layer and the output layer will affect the dynamic processing ability of the dynamic recursive wavelet neural network model. The two association layers belong to the state feedback inside the dynamic recursive wavelet neural network model, forming the dynamic memory performance unique to the recursiveness of the dynamic recursive wavelet neural network model, improving the accuracy and dynamic performance of the dynamic recursive wavelet neural network model; a group of connection weights are added between the first association layer nodes and the output layer nodes of the dynamic recursive wavelet neural network model to enhance the dynamic approximation ability of the dynamic recursive wavelet neural network model and improve the accuracy of controlling the motor speed.

[0076] 6. Design of NARX neural network controller

[0077] The sum of the output of the hybrid neural network controller and the output of the NARX neural network controller is used as the input of the NARX neural network controller and the TDNN neural network controller respectively. The output of the TDNN neural network controller is used as the corresponding input of the Elman neural network prediction model and the input of the L298 motor drive circuit respectively. The NARX neural network controller is a dynamic recursive neural network with output feedback connection. In terms of topological connection relationship, it is equivalent to a BP neural network with input delay plus a delayed feedback connection from output to input. Its structure consists of an input layer, a delay layer, a hidden layer and an output layer. The input layer nodes are used for signal input, the delay layer nodes are used for time delay of input signals and output feedback signals, the hidden layer nodes use activation functions to perform nonlinear operations on delayed signals, and the output layer nodes are used to linearly weight the hidden layer outputs to obtain the final network output. The output h of the i-th hidden layer node of the NARX neural network controller i for:

[0078]

[0079] The output of the jth output layer node of the NARX neural network controller is o j for:

[0080]

[0081] 7. TDNN neural network controller design

[0082] The sum of the output of the hybrid neural network controller and the output of the NARX neural network controller is used as the input of the NARX neural network controller and the TDNN neural network controller respectively. The output of the TDNN neural network controller is used as the corresponding input of the Elman neural network prediction model and the input of the L298 motor drive circuit respectively. The design process of the TDNN neural network controller refers to the design method of the TDNN neural network model of this patent.

[0083] The following is a brief description of the working principle of the transmission platform:

[0084] First, the mobile vehicle 1 is moved to the required position by using the distance-measuring obstacle-avoiding ultrasonic sensor 104, a pair of steering wheels 102, a pair of driving wheels 103, a fourth stepping motor 105 and a steering gear 109, and the height and angle of the first-level conveyor line 2 are adjusted by using the first-level conveyor line end adjustment mechanism and the first-level conveyor line front end adjustment mechanism. The height of the anti-jump height limit frame 201 is adjusted by using the anti-jump height limit frame height adjustment component to prevent goods from jumping.

[0085] Secondly, put the goods on the primary conveyor line 2, start the roller conveying of the primary conveyor line 2 and the secondary conveyor line 3, and the goods move forward under the action of the primary conveyor line 2 and the secondary conveyor line 3. When the goods need to be moved to another place, set an obstacle at the end of the secondary conveyor line 3, and convey all the goods to the primary conveyor line 2 and the secondary conveyor line 3. Use the ranging obstacle avoidance ultrasonic sensor 104 of the mobile trolley 1, a pair of steering wheels 102, a pair of driving wheels 103, the fourth stepper motor 105 and the servo 109 to move the mobile trolley 1 to another position, and adjust the height and angle of the primary conveyor line 2 through the above-mentioned primary conveyor line end adjustment mechanism and the primary conveyor line front end adjustment mechanism, and then dock with the new equipment.

[0086] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with the technology to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent transformation or modification made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. A mobile intelligent transmission system, characterized in that: The invention comprises a mobile trolley (1), a primary conveyor line (2), and a secondary conveyor line (3), wherein the primary conveyor line (2) is arranged on the mobile trolley (1); the secondary conveyor line (3) is connected to the end of the primary conveyor line (2); the mobile trolley (1) is also provided with an adjustment mechanism for adjusting the height and angle of the primary conveyor line (2) and the secondary conveyor line (3); the goods are sequentially transported to the destination via the primary conveyor line (2) and the secondary conveyor line (3); the adjustment mechanism is respectively provided with a plurality of drive motors; the intelligent transmission system further comprises a plurality of motor speed controllers, the output ends of the motor speed controllers are respectively connected to the corresponding drive motors; the motor speed controller comprises an STM32 single chip microcomputer, an L298 motor drive circuit and a speed sensor, the output of the intelligent controller in the STM32 single chip microcomputer is used as the input of the L298 motor drive circuit, the output of the L298 motor drive circuit is used as the input of the controlled drive motor, and the speed sensor measures the actual speed value of the controlled drive motor; the intelligent controller comprises a DRNN neural network setting model, a DRNN neural network prediction model, a DRNN neural network real-time model, and a TDNN neural network model , Elman neural network prediction model, hybrid neural network controller, NARX neural network controller and TDNN neural network controller; the driving motor speed set value is used as the input of the DRNN neural network setting model, the difference between the driving motor speed set value and the output of the Elman neural network prediction model is used as the input of the DRNN neural network prediction model, the difference between the driving motor speed set value and the speed sensor output value is used as the input of the DRNN neural network real-time model, the outputs of the DRNN neural network setting model, the DRNN neural network prediction model and the DRNN neural network real-time model are used as the corresponding inputs of the TDNN neural network model respectively, the difference between the output of the TDNN neural network model and the speed sensor output value is used as the input of the hybrid neural network controller, the sum of the output of the hybrid neural network controller and the output of the NARX neural network controller is used as the input of the NARX neural network controller and the TDNN neural network controller respectively, the output of the TDNN neural network controller is used as the corresponding input of the Elman neural network prediction model and the input of the L298 motor drive circuit respectively, and the speed sensor output value is used as the corresponding input of the Elman neural network prediction model; The adjustment mechanism includes a first-level conveyor line front end adjustment mechanism and a first-level conveyor line end adjustment mechanism; The front end adjustment mechanism of the primary conveyor line comprises a screw adjustment component A, a disc (216), a pair of push rods (217) and a conveyor line bracket (218); the disc (216) is threadedly connected to one end of the screw adjustment component A; the pair of push rods (217) are symmetrically fixed on the disc (216); the lower surface of the primary conveyor line (2) is rotatably connected to the conveyor line bracket (218); the top end of the push rod (217) is connected to the conveyor line bracket (218); The first-stage conveyor line end adjustment mechanism comprises a screw adjustment assembly B and bosses (212) rotatably connected to both sides of the end of the first-stage conveyor line (2), wherein the screw in the screw adjustment assembly B passes through and is threadedly connected to the bosses (212); The end of the primary conveyor line (2) is also provided with an anti-jump height-limiting mechanism, the anti-jump height-limiting mechanism comprising an anti-jump height-limiting frame (201) and a pair of anti-jump height-limiting brackets (202); the bottom end of the anti-jump height-limiting bracket (202) is fixed to the mobile trolley (1), and the upper end thereof is connected to the end of the primary conveyor line (2); an anti-jump height-limiting frame height adjustment component is also provided on the pair of anti-jump height-limiting brackets (202); The height adjustment component of the anti-jump height limit frame comprises the following structure: grooves (203) are respectively provided on both sides of the outer wall of the anti-jump height limit bracket (202) along the vertical direction thereof; racks (204) are respectively slidably connected in a pair of grooves (203), and the tooth opening direction of the racks (204) faces outwards; first driven gears (205) are meshed on the racks (204); the pair of first driven gears (205) are connected via a first rotating shaft (206); a second driven gear (207) is coaxially arranged on the first rotating shaft (206); the second driven gear (207) is meshed with a first driving gear (208); an output shaft of a first stepper motor (209) is connected to the first driving gear (208); and the first stepper motor (209) is connected to a corresponding motor speed controller; The mobile trolley comprises a mobile platform (101), a pair of steering wheels (102), a pair of driving wheels (103), a driving mechanism and a steering mechanism; the pair of steering wheels (102) and the pair of driving wheels (103) are connected to the mobile platform (101) via shafts; the driving mechanism is drivingly connected to the pair of driving wheels (103); and the steering mechanism is connected to the pair of steering wheels (102); and distance-measuring obstacle-avoidance ultrasonic sensors (104) are respectively provided below the front and rear ends of the mobile platform (101) of the mobile trolley (1).

2. The mobile intelligent transmission system according to claim 1, characterized in that: The hybrid neural network controller consists of a tapped delay line TDL (Taped Delay Line), an Adaline neural network model, a DRNN neural network model and a dynamic recursive wavelet neural network model. The input of the hybrid neural network controller serves as the input of the tapped delay line TDL, and the output of the tapped delay line TDL includes the input signal at the current moment and the input signal delayed by 1 to multiple time steps respectively. The output of the tapped delay line TDL serves as the input of the Adaline neural network model and the DRNN neural network model respectively, and the output of the Adaline neural network model and the DRNN neural network model serves as the corresponding input of the dynamic recursive wavelet neural network model, and the output of the dynamic recursive wavelet neural network model serves as the output of the hybrid neural network controller.

3. The mobile intelligent transmission system according to claim 1, characterized in that: The screw adjustment assembly A and the screw adjustment assembly B have the same structure, and both include a stepper motor and a screw, and the stepper motor is connected to its corresponding motor speed controller.

4. The mobile intelligent transmission system according to claim 1, characterized in that: The front and rear ends of the mobile trolley (1) are also provided with dust covers, the dust covers comprising a reeling rod (110) rotatably connected to the lower surface of the mobile platform (101), an isolation cover (111) reeled on the reeling rod (110), a fifth stepping motor (112) being provided on the reeling rod (110), a receiving slot (113) being provided directly below the reeling rod (110) and facing the isolation cover (111), a pressure sensor (114) being provided at the bottom end of the receiving slot (113), and an infrared sensor (115) being provided on the side wall of the receiving slot (113), both of which are connected to a controller, and an output end of the controller is connected to the fifth stepping motor (112); when the isolation cover (111) is lowered, the distance measurement and obstacle avoidance ultrasonic sensor (104) is located inside the isolation cover (111).

Images