Electric roller control method and control system

Abstract

The invention relates to the technical field of electric roller automatic control, and discloses an electric roller control method and a control system. In a first preset time period, a first starting point and a second starting point are respectively established for the operation stroke of a target electric roller and the operation stroke of a target conveying belt; the PID control coefficient is adjusted in a gradient mode to drive the target electric roller to carry out transportation operation; m groups of stable transportation parameters are obtained according to a fixed time interval, and each group of stable transportation parameters corresponds to the relative position of a target electric roller and a target conveying belt; based on the PID empirical coefficient, the stable transportation parameter and the relative position of the stable transportation parameter, training to obtain a stable transportation model; and executing a stable transportation decision and a jolting transportation decision for the target electric roller, including dynamically adjusting the PID empirical coefficient of the target electric roller in response to the stacking state of the transported goods so as to realize stable transportation or jolting flattening of the transported goods.

Description

Technical Field

[0001] The present invention relates to the technical field of automatic control of electric rollers, and more specifically, to a control method and a control system for an electric roller. Background Art

[0002] The electric roller realizes automatic material handling by directly putting the conveyor belt on the outer ring of the roller and driving the roller. It has been widely used in express delivery conveyor lines due to its compact structure and easy maintenance. However, in a semi-open production environment, foreign matter such as debris and dust are easily embedded in the contact surface between the roller and the conveyor belt, causing the friction characteristics to fluctuate non-constantly with the change of stroke; the traditional control strategy that relies on fixed PID parameters is difficult to compensate for this friction disturbance in real time, which can easily cause unstable transportation of express goods, making it difficult for the scanning device to accurately locate and identify the express goods. Summary of the invention

[0003] The present invention provides a control method and a control system for an electric roller, which solve the technical problems raised in the background technology.

[0004] The present invention provides a method for controlling an electric drum, comprising: Step 1: within a first preset time period, establish a first starting point and a second starting point for the operating stroke of the target electric roller and the target conveyor belt, respectively, and adjust the PID control coefficient by gradient to drive the target electric roller to perform the transport operation; Acquire M groups of stable transport parameters at fixed time intervals, and each group of stable transport parameters corresponds to the relative position of the target electric roller and the target conveyor belt; Step 2, based on the PID empirical coefficient, the stable transport parameter and the relative position of the stable transport parameter, a stable transport model is trained; wherein the stable transport model is used to determine the predicted transport stability at any time within a second preset time period; Step 3, executing a stable transportation decision and a bumpy transportation decision for the target electric drum, including: In response to the stacking state of the transported goods, the PID empirical coefficient of the target electric roller is dynamically adjusted to achieve stable transportation or bumpy flattening of the transported goods; Step 4, synchronously obtain the execution result of the stable transportation decision or the bumpy transportation decision; if the execution result does not meet expectations, update the stable transportation model.

[0005] Furthermore, the transport stability parameters include: the operating stroke displacement distance of the target electric roller and the operating stroke displacement distance of the target conveyor belt.

[0006] Further, relative positions include: Based on the first starting point, each position of the target electric drum in the working stroke is encoded as follows: ,in, Indicates the first starting point, Indicates the operating stroke length of the target electric drum; The relative position with respect to the target conveyor belt includes: Based on the second starting point, the target conveyor belt is encoded at each position of the working stroke as follows: ,in, Indicates the second starting point, Indicates the operating stroke length of the target conveyor belt; For the mth group of stable transport parameters; 1≤m≤M, m is a positive integer; Get the starting position and ending position of the target electric drum's operating stroke corresponding to the mth moment ;in, ; Get the starting position and ending position of the target conveyor belt's operation stroke corresponding to the mth moment ;in, ; Determine the relative position of the mth group of stable transport parameters as and .

[0007] Furthermore, based on the stable transportation parameters and the relative position, a stable transportation model is trained, including: Calculating the transport stability includes: taking the difference between the operating stroke displacement distance of the target electric drum and the operating stroke displacement distance of the target conveyor belt in the rth group of stable transport parameters as the transport stability; The normalized transport stability is used as the sample label; The relative position corresponding to the rth group of stable transport parameters is constructed as a position vector; The PID control coefficients corresponding to the rth group of stable transport parameters are constructed as an operation vector; The position vector and the running vector are concatenated and normalized to obtain sample data; Wherein, 1≤r≤M, r≠m, and r is a positive integer; The position vector represents the position vector constructed by the relative position corresponding to the stable transport parameter; The operation matrix represents the operation vector constructed by the PID control coefficients corresponding to the stable transportation parameters; A stable transportation model is obtained by training based on sample data and sample labels; Among them, the hidden layer of the stable transportation model is constructed based on a one-dimensional convolutional neural network, and the hyperparameters of the hidden layer are updated according to the mean square error loss function.

[0008] Furthermore, the stacking state includes: If the stacking height is greater than or equal to the preset height threshold, the stacking status is abnormal; otherwise, the stacking status is normal.

[0009] Furthermore, within the second preset time period, the PID empirical coefficient of the target electric drum is dynamically adjusted to achieve stable transportation of the transported goods, including: Step 31, obtaining the starting position of the operating stroke of the target electric drum at the kth moment in the second preset time period, and the starting position of the operating stroke of the target conveyor belt; Step 32, initializing and generating S control schemes that meet the constraint conditions; wherein the sth control scheme includes the predicted PID control coefficient at the kth moment, the predicted end position of the operating stroke of the target electric drum at the kth moment, and the predicted end position of the operating stroke of the target conveyor belt at the kth moment; Where, 1≤s≤S, s is a positive integer; The constraints include: The PID control coefficients are all within the corresponding control coefficient adjustment range; The difference between the predicted end position and the starting position of the operating stroke of the target electric drum is less than or equal to the first preset displacement difference; The difference between the predicted end position and the starting position of the operation stroke of the target conveyor belt is less than or equal to the second preset displacement difference; Step 33, based on the starting position and predicted end position of the operating stroke of the target electric drum at the kth moment, and the starting position and predicted end position of the operating stroke of the target conveyor belt at the kth moment, respectively generate a position vector and a reference transport stability of the sth control scheme; Step 34, generating an operation vector of the sth control scheme based on the predicted PID control coefficient; Step 35, concatenating the position vector and the running vector and performing normalization processing to use them as input parameters of the stable transportation model to obtain the predicted transportation stability output by the stable transportation model; Step 36: if the difference between the predicted transport stability and the reference transport stability is less than or equal to the preset difference threshold, retain the sth control scheme; otherwise, delete the sth control scheme; Step 37, repeating steps 34 to 36 until U control schemes are obtained, and proceeding to step 38 or step 39; Step 38, selecting the control scheme with the smallest predicted transportation stability among the U control schemes as the control scheme at the kth moment to achieve stable transportation of the transported goods.

[0010] Furthermore, the PID empirical coefficient of the target electric drum is dynamically adjusted to achieve bump-flattening of the transported goods, including: Step 39, selecting the control scheme with the greatest predicted transportation stability among the U control schemes as the control scheme at the kth moment to achieve bump-smoothing of the transported goods.

[0011] Furthermore, the stable transportation model is updated, including: Synchronously obtain the execution results of the stable transportation decision or the bumpy transportation decision; wherein the execution results include meeting expectations and not meeting expectations; Count the cumulative number of times that do not meet expectations. If the cumulative number reaches the preset cumulative number, a full update of the stable transportation model will be triggered; The full update includes: Collect time series sample data and sample labels at fixed time intervals; Load the sliding time window to extract sample data and sample labels; the end of the sliding time window corresponds to the moment when the full update is triggered; Based on the sample data and sample labels extracted from the sliding time window, a stable transportation model is obtained by fully updating.

[0012] In a second aspect, an electric roller control system is applied to the electric roller control method, comprising: An operation driving module, used to establish a first starting point and a second starting point for the operation stroke of the target electric roller and the target conveyor belt respectively within a first preset time period, and to drive the target electric roller to perform a transportation operation by adjusting a PID control coefficient by gradient; Acquire M groups of stable transport parameters at fixed time intervals, and each group of stable transport parameters corresponds to the relative position of the target electric roller and the target conveyor belt; A data acquisition module is used for PID empirical coefficients, stable transport parameters and relative positions of the stable transport parameters to train a stable transport model; wherein the stable transport model is used to determine the predicted transport stability at any time within a second preset time period; Automatic control module for executing stable transport decisions and bumpy transport decisions for the target drum motor, including: In response to the stacking state of the transported goods, the PID empirical coefficient of the target electric roller is dynamically adjusted to achieve stable transportation or bumpy flattening of the transported goods; The automatic update module synchronously obtains the execution results of the stable transportation decision or the bumpy transportation decision; if the execution result does not meet expectations, the stable transportation model is updated.

[0013] The beneficial effects of the present invention are: the "stable transportation" mode is used to reduce the shaking and slipping of materials during the transportation process, and the "bumpy flattening" mode is used to level the uneven stacking. The synergistic effect of the two significantly shortens the transportation cycle, reduces downtime and material loss, thereby maximizing the overall transportation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a module diagram of a method for controlling an electric roller of the present invention; Figure 2 It is a flow chart of a motorized roller control system of the present invention. DETAILED DESCRIPTION

[0015] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0016] like Figure 1 and Figure 2 As shown, a method for controlling an electric drum includes: Step 1: within a first preset time period, establish a first starting point and a second starting point for the operating stroke of the target electric roller and the target conveyor belt, respectively, and adjust the PID control coefficient by gradient to drive the target electric roller to perform the transport operation; Acquire M groups of stable transport parameters at fixed time intervals, and each group of stable transport parameters corresponds to the relative position of the target electric roller and the target conveyor belt; Step 2, based on the PID empirical coefficient, the stable transport parameter and the relative position of the stable transport parameter, a stable transport model is trained; wherein the stable transport model is used to determine the predicted transport stability at any time within a second preset time period; Step 3, executing a stable transportation decision and a bumpy transportation decision for the target electric drum, including: In response to the stacking state of the transported goods, the PID empirical coefficient of the target electric roller is dynamically adjusted to achieve stable transportation or bumpy flattening of the transported goods; Step 4, synchronously obtain the execution result of the stable transportation decision or the bumpy transportation decision; if the execution result does not meet expectations, update the stable transportation model.

[0017] In one embodiment of the present invention, the adjustment range of the PID control coefficient is confirmed by an expert, and the gradient adjustment is performed based on the adjustment range of the PID control coefficient confirmed by the expert.

[0018] Specifically, in the control system of the electric roller, the PID control coefficient (proportional P, integral I, differential D) is the core parameter of the PID controller, and the PID controller is the key tool to achieve precise speed regulation and stable transportation of the electric roller.

[0019] Specifically, the stable transportation decision corresponds to the stable transportation of the transported goods, and the bumpy transportation decision corresponds to the bumpy flattening of the transported goods.

[0020] For example, for an assembly line that transports small express parcels, it is driven by electric rollers and transported by conveyor belts. After the height of the express parcel stack is determined to have exceeded the height limit through a height measuring device (such as an infrared sensor), the bumpy transportation decision is initiated to allow the express parcel to slide and rub, thereby flattening the express parcel to facilitate scanning by the barcode scanning device to avoid problems such as missed scanning and missed inspection. When the height of the express parcel does not exceed the height limit, the stable transportation decision is initiated to stabilize the position of the express parcel to facilitate positioning and scanning by the barcode scanning device.

[0021] In one embodiment of the present invention, the transport stability parameter includes: an operating stroke displacement distance of a target electric drum and an operating stroke displacement distance of a target conveyor belt.

[0022] In detail, the displacement distance is obtained by measuring, for example: An RDIF tag is set on the surface of the conveyor belt, and a distance measuring device is set on the other side. When the conveyor belt is in operation, the distance measuring device measures the displacement distance of the conveyor belt. At the same time, when the distance measuring device senses the RDIF tag, the distance measuring device is reset to zero again. Similarly, the corresponding electric roller measurement.

[0023] In one embodiment of the present invention, the relative position includes: Based on the first starting point, each position of the target electric drum in the working stroke is encoded as follows: ,in, Indicates the first starting point, Indicates the operating stroke length of the target electric drum; The relative position with respect to the target conveyor belt includes: Based on the second starting point, the target conveyor belt is encoded at each position of the working stroke as follows: ,in, Indicates the second starting point, Indicates the operating stroke length of the target conveyor belt; For the mth group of stable transport parameters; 1≤m≤M, m is a positive integer; Get the starting position and ending position of the target electric drum's operating stroke corresponding to the mth moment ;in, ; Get the starting position and ending position of the target conveyor belt's operation stroke corresponding to the mth moment ;in, ; Determine the relative position of the mth group of stable transport parameters as and .

[0024] In one embodiment of the present invention, the target operating stroke of the electric roller is the circumference of the outer side of the electric roller, and the target operating stroke of the conveyor belt is the length of the conveyor belt.

[0025] In one embodiment of the present invention, data of the distance measuring device is acquired at fixed time intervals as the starting position and the ending position at the corresponding moment.

[0026] In one embodiment of the present invention, the working stroke represents the running direction and running path of the electric drum, and the running direction and running path of the conveyor belt.

[0027] In one embodiment of the present invention, based on the stable transport parameters and the relative position, a stable transport model is obtained by training, including: Calculating the transport stability includes: taking the difference between the operating stroke displacement distance of the target electric drum and the operating stroke displacement distance of the target conveyor belt in the rth group of stable transport parameters as the transport stability; The normalized transport stability is used as the sample label; The relative position corresponding to the rth group of stable transport parameters is constructed as a position vector; The PID control coefficients corresponding to the rth group of stable transport parameters are constructed as an operation vector; The position vector and the running vector are concatenated and normalized to obtain sample data; Wherein, 1≤r≤M, r≠m, and r is a positive integer; The position vector represents the position vector constructed by the relative position corresponding to the stable transport parameter; The operation matrix represents the operation vector constructed by the PID control coefficients corresponding to the stable transportation parameters; A stable transportation model is obtained by training based on sample data and sample labels; Among them, the hidden layer of the stable transportation model is constructed based on a one-dimensional convolutional neural network, and the hyperparameters of the hidden layer are updated according to the mean square error loss function.

[0028] In detail, the smaller the transport stability (tending to 0), the better the displacement synchronization between the roller and the conveyor belt, and the less likely the express package is to slip or jitter due to speed difference; conversely, the greater the transport stability, the higher the risk of slipping of the express package.

[0029] For example, in the rth group of data, the roller displacement is 0.5 meters, the conveyor belt displacement is 0.6, and the transportation stability is 0.1 meters. At this time, the package may move backward due to the faster conveyor belt, resulting in a deviation in the code scanning positioning.

[0030] Specifically, in order to eliminate the dimensional differences in transport stability under different working conditions, it is necessary to normalize it (normalize the maximum and minimum values) to a dimensionless value in the range of 0 to 1.

[0031] Relative position is the absolute position of the roller and the conveyor belt normalized to a dimensionless value between 0 and 1. The position vector is a four-dimensional vector. In a certain set of data, the roller runs from 0.2 to 0.5, and the conveyor belt runs from 0.3 to 0.6. The position vector is The operation vector is a combination of the three PID control parameters and is a three-dimensional vector. The position vector and the operation vector are combined into an input that the model can process.

[0032] The core of the stable transport model is to learn the mapping relationship between sample data and sample labels through a one-dimensional convolutional neural network. The typical one-dimensional convolutional neural network structure includes: Input layer: receives a seven-dimensional normalized feature vector. Convolutional layer 1 uses 32 filters, kernel size 3 (covering 3 continuous features), and activation function ReLU to extract local features. Convolutional layer 2 uses 64 filters, kernel size 3, and activation function ReLU to further extract features. The global average pooling layer compresses the output of the convolutional layer into a one-dimensional vector to reduce the computational complexity. The fully connected layer outputs 1 neuron to predict the normalized transport stability. The mean square error (MSE) loss function is used to calculate the deviation between the predicted value and the true value. The weights of the convolutional layer and the fully connected layer are updated through the back-propagation algorithm to minimize the loss value so that the model output is as close to the sample label as possible.

[0033] In one embodiment of the present invention, the stacking state includes: If the stacking height is greater than or equal to the preset height threshold, the stacking status is abnormal; otherwise, the stacking status is normal.

[0034] In one embodiment of the present invention, the preset height threshold is determined by express delivery industry staff.

[0035] In one embodiment of the present invention, dynamically adjusting the PID empirical coefficient of the target electric drum within the second preset time period to achieve stable transportation of the transported goods includes: Step 31, obtaining the starting position of the operating stroke of the target electric drum at the kth moment in the second preset time period, and the starting position of the operating stroke of the target conveyor belt; Step 32, initializing and generating S control schemes that meet the constraint conditions; wherein the sth control scheme includes the predicted PID control coefficient at the kth moment, the predicted end position of the operating stroke of the target electric drum at the kth moment, and the predicted end position of the operating stroke of the target conveyor belt at the kth moment; Where, 1≤s≤S, s is a positive integer; The constraints include: The PID control coefficients are all within the corresponding control coefficient adjustment range; The difference between the predicted end position and the starting position of the operating stroke of the target electric drum is less than or equal to the first preset displacement difference; The difference between the predicted end position and the starting position of the operation stroke of the target conveyor belt is less than or equal to the second preset displacement difference; Step 33, based on the starting position and predicted end position of the operating stroke of the target electric drum at the kth moment, and the starting position and predicted end position of the operating stroke of the target conveyor belt at the kth moment, respectively generate a position vector and a reference transport stability of the sth control scheme; Step 34, generating an operation vector of the sth control scheme based on the predicted PID control coefficient; Step 35, concatenating the position vector and the running vector and performing normalization processing to use them as input parameters of the stable transportation model to obtain the predicted transportation stability output by the stable transportation model; Step 36: if the difference between the predicted transport stability and the reference transport stability is less than or equal to the preset difference threshold, retain the sth control scheme; otherwise, delete the sth control scheme; Step 37, repeating steps 34 to 36 until U control schemes are obtained, and proceeding to step 38 or step 39; Step 38, selecting the control scheme with the smallest predicted transportation stability among the U control schemes as the control scheme at the kth moment to achieve stable transportation of the transported goods.

[0036] In one embodiment of the present invention, dynamically adjusting the PID empirical coefficient of the target electric drum to achieve bumpy flattening of transported goods includes: Step 39, selecting the control scheme with the greatest predicted transportation stability among the U control schemes as the control scheme at the kth moment to achieve bump-smoothing of the transported goods.

[0037] In one embodiment of the present invention, U<S, and U is a positive integer.

[0038] In one embodiment of the present invention, all U executable solutions at the kth moment are determined based on an optimization algorithm, specifically including: First, S control schemes that meet the constraints are randomly generated, including: PID control coefficients and the displacement distance of the electric drum and the displacement distance of the conveyor belt.

[0039] If the difference between the transport stability corresponding to the random control scheme and the transport stability output by the model is small (less than or equal to the threshold), it means that the scheme is feasible. Otherwise, it means that the random scheme is not feasible.

[0040] Then, until U feasible control schemes are obtained, the stacking state at the corresponding moment is responded to. If the stacking state is too high, the control scheme with the maximum transport stability is selected for execution, otherwise the control scheme with the minimum transport stability is selected for execution, to correspond to the bump flattening decision and the stable transport decision respectively.

[0041] In one embodiment of the present invention, updating the stable transportation model includes: Synchronously obtain the execution results of the stable transportation decision or the bumpy transportation decision; wherein the execution results include meeting expectations and not meeting expectations; Count the cumulative number of times that do not meet expectations. If the cumulative number reaches the preset cumulative number, a full update of the stable transportation model will be triggered; The full update includes: Collect time series sample data and sample labels at fixed time intervals; Load the sliding time window to extract sample data and sample labels; the end of the sliding time window corresponds to the moment when the full update is triggered; Based on the sample data and sample labels extracted from the sliding time window, a stable transportation model is obtained by fully updating.

[0042] In one embodiment of the present invention, the expected result corresponding to the stable transport decision is stable transport, while the expected result corresponding to the bumpy flattening decision is bumpy flattening. Since the implementation of these two transport decisions is based on transport stability. However, transport stability will change over time, resulting in results that do not meet expectations. Therefore, the stable transport model needs to be retrained based on the latest data at regular intervals.

[0043] An electric roller control system, applied to the electric roller control method, comprises: An operation driving module, used to establish a first starting point and a second starting point for the operation stroke of the target electric roller and the target conveyor belt respectively within a first preset time period, and to drive the target electric roller to perform a transportation operation by adjusting a PID control coefficient by gradient; Acquire M groups of stable transport parameters at fixed time intervals, and each group of stable transport parameters corresponds to the relative position of the target electric roller and the target conveyor belt; A data acquisition module is used for PID empirical coefficients, stable transport parameters and relative positions of the stable transport parameters to train a stable transport model; wherein the stable transport model is used to determine the predicted transport stability at any time within a second preset time period; Automatic control module for executing stable transport decisions and bumpy transport decisions for the target drum motor, including: In response to the stacking state of the transported goods, the PID empirical coefficient of the target electric roller is dynamically adjusted to achieve stable transportation or bumpy flattening of the transported goods; The automatic update module synchronously obtains the execution results of the stable transportation decision or the bumpy transportation decision; if the execution result does not meet expectations, the stable transportation model is updated.

[0044] It should be noted that the interval and threshold size are set for the convenience of comparison, where the size of the threshold depends on the amount of sample data and the number of bases set by technicians in this field for each group of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. In addition, the above formulas are all calculations of removing dimensions and taking their values. The formulas are all obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0045] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. A method for controlling an electric drum, characterized in that: include: Step 1: within a first preset time period, establish a first starting point and a second starting point for the operating stroke of the target electric roller and the target conveyor belt, respectively, and adjust the PID control coefficient by gradient to drive the target electric roller to perform the transport operation; Acquire M groups of stable transport parameters at fixed time intervals, and each group of stable transport parameters corresponds to the relative position of the target electric roller and the target conveyor belt; Step 2, based on the PID empirical coefficient, the stable transport parameter and the relative position of the stable transport parameter, a stable transport model is trained; wherein the stable transport model is used to determine the predicted transport stability at any time within a second preset time period; Step 3, executing a stable transportation decision and a bumpy transportation decision for the target electric drum, including: In response to the stacking state of the transported goods, the PID empirical coefficient of the target electric roller is dynamically adjusted to achieve stable transportation or bumpy flattening of the transported goods; Step 4, synchronously obtain the execution result of the stable transportation decision or the bumpy transportation decision; if the execution result does not meet expectations, update the stable transportation model.

2. A method for controlling an electric drum according to claim 1, characterized in that: Transport stability parameters include: operating travel displacement distance of the target electric drum and operating travel displacement distance of the target conveyor belt.

3. A method for controlling an electric drum according to claim 2, characterized in that: Relative position, including: Based on the first starting point, each position of the target electric drum in the working stroke is encoded as follows: ,in, Indicates the first starting point, Indicates the operating stroke length of the target electric drum; The relative position with respect to the target conveyor belt includes: Based on the second starting point, the target conveyor belt is encoded at each position of the working stroke as follows: ,in, Indicates the second starting point, Indicates the operating stroke length of the target conveyor belt; For the mth group of stable transport parameters; 1≤m≤M, m is a positive integer; Get the starting position and ending position of the target electric drum's operating stroke corresponding to the mth moment ;in, ; Get the starting position and ending position of the target conveyor belt's operation stroke corresponding to the mth moment ;in, ; Determine the relative position of the mth group of stable transport parameters as and .

4. A method for controlling an electric drum according to claim 3, characterized in that: Based on the stable transport parameters and relative positions, a stable transport model is trained, including: Calculating the transport stability includes: taking the difference between the operating stroke displacement distance of the target electric drum and the operating stroke displacement distance of the target conveyor belt in the rth group of stable transport parameters as the transport stability; The normalized transport stability is used as the sample label; The relative position corresponding to the rth group of stable transport parameters is constructed as a position vector; The PID control coefficients corresponding to the rth group of stable transport parameters are constructed as an operation vector; The position vector and the running vector are concatenated and normalized to obtain sample data; Wherein, 1≤r≤M, r≠m, and r is a positive integer; The position vector represents the position vector constructed by the relative position corresponding to the stable transport parameter; The operation matrix represents the operation vector constructed by the PID control coefficients corresponding to the stable transportation parameters; A stable transportation model is obtained by training based on sample data and sample labels; Among them, the hidden layer of the stable transportation model is constructed based on a one-dimensional convolutional neural network, and the hyperparameters of the hidden layer are updated according to the mean square error loss function.

5. A method for controlling an electric drum according to claim 4, characterized in that: Stack status, including: If the stacking height is greater than or equal to the preset height threshold, the stacking status is abnormal; otherwise, the stacking status is normal.

6. A method for controlling an electric drum according to claim 5, characterized in that: During the second preset time period, the PID empirical coefficient of the target electric drum is dynamically adjusted to achieve stable transportation of the transported goods, including: Step 31, obtaining the starting position of the operating stroke of the target electric drum at the kth moment in the second preset time period, and the starting position of the operating stroke of the target conveyor belt; Step 32, initializing and generating S control schemes that meet the constraint conditions; wherein the sth control scheme includes the predicted PID control coefficient at the kth moment, the predicted end position of the operating stroke of the target electric drum at the kth moment, and the predicted end position of the operating stroke of the target conveyor belt at the kth moment; Where, 1≤s≤S, s is a positive integer; The constraints include: The PID control coefficients are all within the corresponding control coefficient adjustment range; The difference between the predicted end position and the starting position of the operating stroke of the target electric drum is less than or equal to the first preset displacement difference; The difference between the predicted end position and the starting position of the operation stroke of the target conveyor belt is less than or equal to the second preset displacement difference; Step 33, based on the starting position and predicted end position of the operating stroke of the target electric drum at the kth moment, and the starting position and predicted end position of the operating stroke of the target conveyor belt at the kth moment, respectively generate a position vector and a reference transport stability of the sth control scheme; Step 34, generating an operation vector of the sth control scheme based on the predicted PID control coefficient; Step 35, concatenating the position vector and the running vector and performing normalization processing to use them as input parameters of the stable transportation model to obtain the predicted transportation stability output by the stable transportation model; Step 36: if the difference between the predicted transport stability and the reference transport stability is less than or equal to the preset difference threshold, retain the sth control scheme; otherwise, delete the sth control scheme; Step 37, repeating steps 34 to 36 until U control schemes are obtained, and proceeding to step 38 or step 39; Step 38, selecting the control scheme with the smallest predicted transportation stability among the U control schemes as the control scheme at the kth moment to achieve stable transportation of the transported goods.

7. A method for controlling an electric drum according to claim 6, characterized in that: Dynamically adjust the PID empirical coefficient of the target electric drum to achieve bump-flattening of transported goods, including: Step 39, selecting the control scheme with the greatest predicted transportation stability among the U control schemes as the control scheme at the kth moment to achieve bump-smoothing of the transported goods.

8. A method for controlling an electric drum according to claim 7, characterized in that: Updated stable transport model, including: Synchronously obtain the execution results of the stable transportation decision or the bumpy transportation decision; wherein the execution results include meeting expectations and not meeting expectations; Count the cumulative number of times that do not meet expectations. If the cumulative number reaches the preset cumulative number, a full update of the stable transportation model will be triggered; The full update includes: Collect time series sample data and sample labels at fixed time intervals; Load the sliding time window to extract sample data and sample labels; the end of the sliding time window corresponds to the moment when the full update is triggered; Based on the sample data and sample labels extracted from the sliding time window, a stable transportation model is obtained by fully updating.

9. An electric roller control system, applied to an electric roller control method according to any one of claims 1 to 8, characterized in that: include: An operation driving module, used to establish a first starting point and a second starting point for the operation stroke of the target electric roller and the target conveyor belt respectively within a first preset time period, and to drive the target electric roller to perform a transportation operation by adjusting a PID control coefficient by gradient; Acquire M groups of stable transport parameters at fixed time intervals, and each group of stable transport parameters corresponds to the relative position of the target electric roller and the target conveyor belt; A data acquisition module is used for PID empirical coefficients, stable transport parameters and relative positions of the stable transport parameters to train a stable transport model; wherein the stable transport model is used to determine the predicted transport stability at any time within a second preset time period; Automatic control module for executing stable transport decisions and bumpy transport decisions for the target drum motor, including: In response to the stacking state of the transported goods, the PID empirical coefficient of the target electric roller is dynamically adjusted to achieve stable transportation or bumpy flattening of the transported goods; The automatic update module synchronously obtains the execution results of the stable transportation decision or the bumpy transportation decision; if the execution result does not meet expectations, the stable transportation model is updated.

Images