Multi-source error cooperative compensation method and system based on cold-rolled sheet online detection

Abstract

The invention relates to a multi-source error cooperative compensation method and system based on cold-rolled sheet online detection, and belongs to the technical field of error monitoring. The method comprises the following steps: acquiring multi-dimensional data of a cold-rolled sheet in real time, and processing the multi-dimensional data of the cold-rolled sheet through a deep space-time error coupling network to generate a cold-rolling error propagation chain type relation graph; establishing a cold rolling error dynamic weight distribution mechanism according to the cold rolling error propagation chain type relation graph; calculating through the cold rolling error dynamic weight distribution mechanism to obtain a rolling multi-dimensional cooperative compensation coefficient; and outputting a multi-source cold rolling error cooperative compensation amount according to the rolling multi-dimensional cooperative compensation coefficient, thereby responding to the change of working conditions in real time, automatically adjusting the rolling multi-dimensional cooperative compensation priority, and realizing multi-source error cooperative compensation.

Description

Technical Field

[0001] The present invention belongs to the technical field of error monitoring, and particularly relates to a multi-source error collaborative compensation method and system based on on-line detection of cold-rolled thin plates. Background Technique

[0002] As a key material in the fields of automobiles, household appliances, precision instruments, etc., cold-rolled thin plates have extremely high requirements for thickness accuracy and surface quality. The existing on-line detection technology for cold-rolled thin plates mainly relies on equipment such as laser thickness gauges, X-ray residual stress detectors, and multi-channel eddy current sensors, combined with PID control and statistical process control.

[0003] However, there is strong non-linear coupling among multi-source errors such as mechanical vibration, thermal deformation, and material microtexture evolution during the rolling process. The traditional single-variable compensation model cannot decouple the dynamic correlation. The existing compensation algorithm based on FEM has a long calculation period and is difficult to meet the real-time requirement. Moreover, the laser thickness gauge is affected by spot noise caused by vibration and has a temperature drift, resulting in large fluctuations in thickness detection errors. Summary of the Invention

[0004] In order to solve the above problems existing in the prior art, the present invention proposes a multi-source error collaborative compensation method and system based on on-line detection of cold-rolled thin plates.

[0005] The object of the present invention can be achieved by the following technical solutions:

[0006] A multi-source error collaborative compensation method based on on-line detection of cold-rolled thin plates includes:

[0007] S1: Real-time collect multi-dimensional data of cold-rolled thin plates, and process the multi-dimensional data of cold-rolled thin plates through a deep spatio-temporal error coupling network to generate a cold-rolled error propagation chain relationship map;

[0008] S2: Establish a cold-rolled error dynamic weight distribution mechanism according to the cold-rolled error propagation chain relationship map;

[0009] S3: Calculate the rolling multi-dimensional collaborative compensation coefficient through the cold-rolled error dynamic weight distribution mechanism;

[0010] S4: Output the multi-source cold-rolled error collaborative compensation amount according to the rolling multi-dimensional collaborative compensation coefficient.

[0011] Preferably, the generation process of the cold-rolled error propagation chain relationship map in step S1 is:

[0012] Receive the multi-dimensional data of the cold-rolled thin sheet through the deep spatio-temporal error coupling network, extract the spatial distribution pattern through the spatial feature extraction unit, reorganize the features through the temporal feature extraction unit for bidirectional LSTM processing to capture the dynamic characteristics related to the rolling speed, embed the rolling force equation through the physical constraint embedding unit to correct the feature distribution for physical law constraint, and perform topological relationship modeling through the graph convolution propagation unit to output and generate a cold-rolled error propagation chain relationship graph.

[0013] Preferably, the structure of the cold-rolled error propagation chain relationship graph in step S1 is a dynamic topological structure, including equipment nodes, process nodes, error nodes, and environment nodes. The nodes are connected by directed edges to represent the error propagation path, and the weight of the edge quantifies the deep spatio-temporal error coupling strength.

[0014] Preferably, the cold-rolled error dynamic weight allocation mechanism in step S2 is as follows:

[0015] S201: Obtain the cold-rolled thin sheet error data according to the cold-rolled error propagation chain relationship graph;

[0016] S202: Calculate the initial weight of the cold-rolled thin sheet error source according to the cold-rolled thin sheet error data through the error source information entropy;

[0017] S203: Optimize the initial weight of the cold-rolled thin sheet error source through random search to obtain the weight of the cold-rolled thin sheet error source.

[0018] Preferably, the mathematical expression for calculating the error source information entropy in step S202 is:

[0019]

[0020] where E i is the information entropy of the i-th error source, m is the total number of samples, k is the sample index, X ik is the normalized value of the k-th sample of the i-th error source, and p ik is the proportion of the k-th sample of the i-th error source.

[0021] Preferably, the calculation process of the rolling multi-dimensional collaborative compensation coefficient in step S3 is as follows:

[0022] S301: Obtain the error sensitivity by processing the weight of the cold-rolled thin sheet error source through error sensitivity calculation;

[0023] S302: Output the rolling multi-dimensional collaborative compensation coefficient by processing the error sensitivity through compensation priority allocation.

[0024] Preferably, the generation process of the multi-source cold-rolled error collaborative compensation amount in step S4 is as follows:

[0025] S401: Process the rolling multi - dimensional collaborative compensation coefficient through sub - execution allocation to obtain a cold - rolling error dynamic compensation allocation matrix;

[0026] S402: Process the cold - rolling error dynamic compensation allocation matrix through global correction constraints to obtain a multi - source cold - rolling error collaborative compensation amount.

[0027] Preferably, the mathematical expression of the cold - rolling error dynamic compensation allocation matrix in step S401 is:

[0028] M = W d *C + B,

[0029] where M is the cold - rolling error dynamic compensation allocation matrix, W d is the actuator dynamic weight matrix, B is the actuator physical boundary constraint, and C is the rolling multi - dimensional collaborative compensation coefficient vector.

[0030] Preferably, the process of global correction constraints in step S402 is as follows:

[0031] S402 - 1: Generate a penalty - enhanced objective function through adaptive penalty function embedding according to the cold - rolling error dynamic compensation allocation matrix;

[0032] S402 - 2: Reconstruct the slack variables according to the penalty - enhanced objective function to obtain a slack multi - constraint optimization model;

[0033] S402 - 3: Solve the cold - rolling error dynamic compensation allocation matrix through multi - objective gradient descent according to the slack multi - constraint optimization model to obtain a cold - rolling error dynamic compensation allocation correction matrix;

[0034] S402 - 4: Generate a multi - source cold - rolling error collaborative compensation amount by processing the cold - rolling error dynamic compensation allocation correction matrix through actuator physical limit - amplitude.

[0035] A multi - source error collaborative compensation system based on on - line detection of cold - rolled thin plates, which is applied to the above - mentioned multi - source error collaborative compensation method, includes a depth - time - space error coupling module, a cold - rolling error dynamic weight allocation module, a collaborative compensation coefficient calculation module, and a cold - rolling error collaborative compensation module;

[0036] The depth - time - space error coupling module is used to collect multi - dimensional data of cold - rolled thin plates in real - time and generate a cold - rolling error propagation chain - type relationship graph through a depth - time - space error coupling network;

[0037] The cold - rolling error dynamic weight allocation module is used to establish a cold - rolling error dynamic weight allocation mechanism according to the cold - rolling error propagation chain - type relationship graph;

[0038] The collaborative compensation coefficient calculation module is used to calculate the rolling multi-dimensional collaborative compensation coefficient through the cold rolling error dynamic weight allocation mechanism;

[0039] The cold rolling error collaborative compensation module is used to output the multi-source cold rolling error collaborative compensation amount according to the rolling multi-dimensional collaborative compensation coefficient.

[0040] The beneficial effects of the present invention are as follows:

[0041] (1) Explicitly analyze the spatio-temporal propagation path of multi-source errors in the rolling process through the deep spatio-temporal error coupling network, realize the accurate positioning and collaborative decoupling of error sources, and quantify the contribution degrees of different error sources by constructing an error propagation chain diagram.

[0042] (2) Through the dynamic weight allocation mechanism, automatically adjust the priority of rolling multi-dimensional collaborative compensation, can respond to the working condition changes in real time, and realize the multi-source error collaborative compensation for on-line detection of cold-rolled thin plates.

[0043] (3) Obtain the dynamic compensation allocation matrix by distributing and processing the rolling multi-dimensional collaborative compensation coefficient; obtain the multi-source cold rolling error collaborative compensation amount by globally correcting and constraining the dynamic compensation allocation matrix, dynamically allocate the rolling multi-dimensional collaborative compensation coefficient according to the physical characteristics and process constraints of the actuators, construct a cold rolling error dynamic compensation allocation matrix, and solve the resource competition problem in multi-actuator coupling control. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings.

[0045] Figure 1 It is a schematic flow chart of a multi-source error collaborative compensation method based on on-line detection of cold-rolled thin plates of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] In order to further elaborate the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following will, with reference to the accompanying drawings and preferred embodiments, describe in detail the specific embodiments, structures, features and their effects according to the present invention.

[0047] Please refer to Figure 1 , a multi-source error collaborative compensation method based on on-line detection of cold-rolled thin plates, including:

[0048] S1: Real-time collect multi-dimensional data of cold-rolled thin plates, and generate a cold rolling error propagation chain diagram by processing the multi-dimensional data of cold-rolled thin plates through a deep spatio-temporal error coupling network;

[0049] S2: Establish a cold rolling error dynamic weight allocation mechanism according to the cold rolling error propagation chain diagram;

[0050] S3: Calculate the rolling multi-dimensional collaborative compensation coefficient through the cold rolling error dynamic weight distribution mechanism;

[0051] S4: Output the multi-source cold rolling error collaborative compensation amount according to the rolling multi-dimensional collaborative compensation coefficient.

[0052] Embodiment 1

[0053] In this embodiment, the real-time acquisition of multi-dimensional data of cold-rolled thin plates, and the generation of a cold rolling error propagation chain relationship map by processing the multi-dimensional data of cold-rolled thin plates through a deep spatio-temporal error coupling network are specifically implemented through the following steps:

[0054] S101: The multi-dimensional data of cold-rolled thin plates includes strip shape stress distribution data, roll gap dynamic gap data, roll thermal expansion coefficient data, and tension fluctuation value data;

[0055] S101-1: Obtain the strip shape stress distribution data through a multi-frequency eddy current array sensor;

[0056] S101-2: Obtain the roll gap dynamic gap data through a dual-frequency laser interferometer;

[0057] S101-3: Obtain the roll thermal expansion coefficient data through an infrared thermal imager and a fiber Bragg grating sensor;

[0058] S101-4: Obtain the tension fluctuation value data through a laser Doppler velocimeter;

[0059] It should be noted that the multi-dimensional data of the cold-rolled thin plates collected in real time needs to be preprocessed, and the data preprocessing includes time synchronization, spatial registration, and anomaly detection.

[0060] The time synchronization is through the FPGA hardware clock synchronization to ensure that the time stamp deviation of multi-sensor data is less than the preset time threshold, and at the same time, cubic spline interpolation is performed on the low-frequency sensor data to match the time series of high-frequency data.

[0061] The spatial registration establishes a rolling line coordinate system, maps the data of each sensor to a unified grid through affine transformation, and at the same time uses an RBF neural network to repair the missing data.

[0062] The anomaly detection is based on the sliding window Z-score algorithm, and the data points exceeding the range of ±3σ are marked in real time to trigger the sensor self-check program.

[0063] S102: The deep spatio-temporal error coupling network includes a spatial feature extraction unit, a temporal feature extraction unit, a physical constraint embedding unit, and a graph convolution propagation unit.

[0064] The spatial feature extraction unit constructs a lightweight three-dimensional convolutional network through deep separable convolution, adopts an inverse residual structure to reduce the number of parameters, introduces the SE attention mechanism to enhance the key area features, and outputs the spatial features of the cold-rolled sheet;

[0065] The time series feature extraction unit captures the dynamic characteristics of time-varying parameters such as rolling speed and tension fluctuation through a bidirectional long short-term memory network, establishes a time correlation model of the rolling process, and outputs the time series features of the cold-rolled thin plate;

[0066] The physical constraint embedding unit uses the rolling force balance equation as a hard constraint to ensure that the network prediction complies with the metal plastic deformation law;

[0067] The graph convolution propagation unit models the error transmission relationship between the nodes of the rolling mill equipment by constructing a dynamic adjacency matrix, thereby generating an error propagation chain graph.

[0068] The deep spatiotemporal error coupling network receives the multi-dimensional data of the cold-rolled thin plate, extracts the spatial distribution pattern through a spatial feature extraction unit and a 3D MobileNetV3, reorganizes the features through a temporal feature extraction unit to perform bidirectional LSTM processing to capture dynamic characteristics related to the rolling speed, embeds the rolling force equation through a physical constraint embedding unit to correct the feature distribution to constrain the physical law to ensure that the prediction complies with the material deformation mechanism, models the topological relationship through a graph convolution propagation unit, dynamically constructs the device node adjacency matrix, and generates a cold rolling error propagation chain relationship map through the graph convolution propagation error coupling effect output.

[0069] The structure of the cold rolling error propagation chain relationship map is a dynamic topological structure, including equipment nodes, process nodes, error nodes and environment nodes. The nodes are connected by directed edges to represent the error propagation path (such as "roller thermal deformation → rolling force fluctuation → thickness deviation"). The edge weight quantifies the depth of spatiotemporal error coupling strength (such as roller thermal expansion coefficient). The cold rolling error propagation chain relationship map adopts a hierarchical dynamic adjacency matrix, which integrates the physical connection relationship and the data-driven implicit association. Through the joint modeling of space-time-physical constraints, it explicitly expresses the chain transmission, cross-regional diffusion and multi-target coupling effects of multi-source errors in the rolling process. It has both dynamic evolution capability (real-time update with working conditions) and mechanism interpretability (embedded rolling force balance equation constraints), providing visual causal reasoning support for collaborative compensation.

[0070] In this embodiment, the establishment of a dynamic weight distribution mechanism for cold rolling errors according to the cold rolling error propagation chain relationship graph is specifically implemented by the following steps:

[0071] S201: Obtaining cold-rolled sheet error data according to the cold-rolling error propagation chain relationship map;

[0072] S202: Calculate the initial weight of the cold-rolled sheet error source based on the cold-rolled sheet error data through the error source information entropy;

[0073] The mathematical expression for calculating the error source information entropy is:

[0074]

[0075] where E i is the information entropy of the i-th error source, m is the total number of samples, k is the sample index, and X ik is the standardized value of the k-th sample of the i-th error source, and p ik is the proportion of the k-th sample of the i-th error source;

[0076] The mathematical expression for the initial weight of the cold-rolled sheet error source is:

[0077]

[0078] where E i is the information entropy of the i-th error source, E j is the information entropy of the j-th error source, and w i is the initial weight of the cold-rolled sheet error source of the i-th error source.

[0079] S203: Optimize the initial weight of the cold-rolled sheet error source through random search to obtain the weight of the cold-rolled sheet error source.

[0080] In this embodiment, the rolling multi-dimensional collaborative compensation coefficient calculated through the cold rolling error dynamic weight distribution mechanism is specifically implemented through the following steps:

[0081] S301: Obtain the error sensitivity by processing the weight of the cold-rolled sheet error source through error sensitivity calculation;

[0082] The mathematical expression for the error sensitivity is:

[0083]

[0084] where S is the error sensitivity, w T is the temperature gradient error weight, w x is the mechanical stress error weight, w p is the motion control error weight, ΔT k is the temperature gradient error, R k is the mechanical stress error, M k is the motion control error, T max is the maximum allowable temperature deviation, R yield is the material yield strength, M crit is the critical strain rate.

[0085] S302: Process the error sensitivity output to obtain the rolling multi-dimensional collaborative compensation coefficient through compensation priority allocation.

[0086] When the error sensitivity is greater than or equal to 0.8, the compensation priority is the high-sensitivity level, that is, the error is close to or exceeds the safety threshold, and full-power compensation is required immediately. Skip optimization and directly output the maximum rolling multi-dimensional collaborative compensation coefficient.

[0087] When the error sensitivity is greater than or equal to 0.4 and less than 0.8, the compensation priority is the medium-sensitivity level, that is, the error is significant but controllable, and optimization compensation is required. Call the IPSO algorithm to solve the optimal solution to obtain the rolling multi-dimensional collaborative compensation coefficient.

[0088] When the error sensitivity is less than 0.4, the compensation priority is the low-sensitivity level, that is, the error is within the allowable range, and compensation is carried out after accumulation, and the rolling multi-dimensional collaborative compensation coefficient is output.

[0089] In this embodiment, the output of the multi-source cold rolling error collaborative compensation amount according to the rolling multi-dimensional collaborative compensation coefficient is specifically implemented through the following steps:

[0090] S401: Process the rolling multi-dimensional collaborative compensation coefficient through sub-execution allocation to obtain the cold rolling error dynamic compensation allocation matrix.

[0091] The mathematical expression of the cold rolling error dynamic compensation allocation matrix is:

[0092] M = W d *C + B,

[0093] where M is the cold rolling error dynamic compensation allocation matrix, W d is the actuator dynamic weight matrix, B is the actuator physical boundary constraint, and C is the rolling multi-dimensional collaborative compensation coefficient vector.

[0094] It should be noted that the rolling multi-dimensional collaborative compensation coefficient needs to be dynamically allocated according to the physical characteristics of the actuator and the process constraints.

[0095] S402: Process the cold rolling error dynamic compensation allocation matrix through global correction constraints to obtain the multi-source cold rolling error collaborative compensation amount.

[0096] S402-1: Generate a penalty-enhanced objective function through the embedding of an adaptive penalty function according to the cold rolling error dynamic compensation allocation matrix.

[0097] S402-11: Preset the global constraints of the cold rolling process. The global constraints of the cold rolling process include the rolling force threshold, the allowable range of strip shape flatness, and the energy consumption upper limit.

[0098] S402-12: Traverse each compensation amount in the cold rolling error dynamic compensation distribution matrix. If it exceeds the preset global constraints of the cold rolling process, a penalty-enhanced objective function is obtained by superimposing a penalty term on the original objective function through a non-linear penalty function.

[0099] S402-2: Reconstruct the slack variables according to the penalty-enhanced objective function to obtain a slack multi-constraint optimization model;

[0100] S402-3: Solve the cold rolling error dynamic compensation distribution matrix by multi-objective gradient descent according to the slack multi-constraint optimization model to obtain a cold rolling error dynamic compensation distribution correction matrix;

[0101] S402-4: Generate a multi-source cold rolling error collaborative compensation amount by physically limiting the cold rolling error dynamic compensation distribution correction matrix through an actuator.

[0102] Embodiment 2

[0103] A multi-source error collaborative compensation system based on on-line detection of cold-rolled thin plates includes a depth-time-space error coupling module, a cold rolling error dynamic weight distribution module, a collaborative compensation coefficient calculation module, and a cold rolling error collaborative compensation module;

[0104] The depth-time-space error coupling module is used to collect multi-dimensional data of cold-rolled thin plates in real time and generate a cold rolling error propagation chain relationship map through a depth-time-space error coupling network;

[0105] The cold rolling error dynamic weight distribution module is used to establish a cold rolling error dynamic weight distribution mechanism according to the cold rolling error propagation chain relationship map;

[0106] The collaborative compensation coefficient calculation module is used to calculate a rolling multi-dimensional collaborative compensation coefficient through the cold rolling error dynamic weight distribution mechanism;

[0107] The cold rolling error collaborative compensation module is used to output a multi-source cold rolling error collaborative compensation amount according to the rolling multi-dimensional collaborative compensation coefficient.

[0108] The above is only a preferred embodiment of the present invention, and does not impose any form of limitation on the present invention. Although the present invention has been disclosed above with preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to the above-disclosed technical content within the scope of the technical solution of the present invention to obtain equivalent embodiments with equivalent changes. However, as long as it does not depart from the content of the technical solution of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A multi-source error collaborative compensation method based on on-line detection of cold-rolled thin plates, characterized in that, Including: S1: Collect multi-dimensional data of cold-rolled thin plates in real time, and process the multi-dimensional data of the cold-rolled thin plates through a deep spatio-temporal error coupling network to generate a cold-rolling error propagation chain relationship map; S2: Establish a cold-rolling error dynamic weight allocation mechanism according to the cold-rolling error propagation chain relationship map; S3: Calculate a rolling multi-dimensional collaborative compensation coefficient through the cold-rolling error dynamic weight allocation mechanism; S4: Output a multi-source cold-rolling error collaborative compensation amount according to the rolling multi-dimensional collaborative compensation coefficient.

2. The multi-source error collaborative compensation method according to claim 1, wherein The generation process of the cold-rolling error propagation chain relationship map in step S1 is as follows: Receive the multi-dimensional data of the cold-rolled thin plates through a deep spatio-temporal error coupling network, extract the spatial distribution pattern through a spatial feature extraction unit, recombine features through a time series feature extraction unit for bidirectional LSTM processing to capture dynamic characteristics related to rolling speed, embed the rolling force equation through a physical constraint embedding unit to correct the feature distribution for physical law constraint, and perform topological relationship modeling through a graph convolution propagation unit to output and generate a cold-rolling error propagation chain relationship map.

3. The multi-source error collaborative compensation method according to claim 1, characterized in that The structure of the cold-rolling error propagation chain relationship map in step S1 is a dynamic topological structure, including equipment nodes, process nodes, error nodes and environment nodes. The nodes are connected by directed edges to represent the error propagation path, and the weight of the edge quantifies the depth spatio-temporal error coupling strength.

4. The multi-source error collaborative compensation method according to claim 1, characterized in that The cold-rolling error dynamic weight allocation mechanism in step S2 is as follows: S201: Obtain cold-rolled thin plate error data according to the cold-rolling error propagation chain relationship map; S202: Calculate the initial weight of the cold-rolled thin plate error source through error source information entropy according to the cold-rolled thin plate error data; S203: Optimize the initial weight of the cold-rolled thin plate error source through random search to obtain the weight of the cold-rolled thin plate error source.

5. The multi-source error collaborative compensation method according to claim 4, characterized in that The mathematical expression for calculating the error source information entropy in step S202 is: Among them, E i is the information entropy of the i-th error source, m is the total number of samples, k is the sample index, and X ik is the standardized value of the k-th sample of the i-th error source, and p ik is the proportion of the k-th sample of the i-th error source.

6. The multi-source error collaborative compensation method according to claim 1, wherein The calculation process of the rolling multi-dimensional collaborative compensation coefficient in step S3 is as follows: S301: Obtain the error sensitivity by processing the weight of the cold-rolled thin plate error source through error sensitivity calculation; S302: Output a rolling multi-dimensional collaborative compensation coefficient by processing the error sensitivity through compensation priority allocation.

7. The multi-source error collaborative compensation method according to claim 1, characterized in that The generation process of the multi-source cold-rolling error collaborative compensation amount in step S4 is as follows: S401: Obtain a cold-rolling error dynamic compensation allocation matrix by processing the rolling multi-dimensional collaborative compensation coefficient through sub-execution allocation; S402: Obtain a multi-source cold-rolling error collaborative compensation amount by processing the cold-rolling error dynamic compensation allocation matrix through global correction constraint.

8. The multi-source error collaborative compensation method according to claim 7, wherein The mathematical expression for the cold-rolling error dynamic compensation allocation matrix in step S401 is: M = W d * C + B, Among them, M is the dynamic compensation distribution matrix for cold rolling errors, W d is the dynamic weight matrix of the actuator, B is the physical boundary constraint of the actuator, and C is the multi-dimensional collaborative compensation coefficient vector for rolling.

9. The multi-source error collaborative compensation method according to claim 7, wherein The process of global correction constraint processing in step S402 is as follows: S402-1: Generate a penalty-enhanced objective function by embedding an adaptive penalty function according to the cold-rolling error dynamic compensation allocation matrix; S402-2: Reconstruct the slack variable according to the penalty-enhanced objective function to obtain a slack multi-constraint optimization model; S402-3: Solve the cold-rolling error dynamic compensation allocation matrix through multi-objective gradient descent according to the slack multi-constraint optimization model to obtain a cold-rolling error dynamic compensation allocation correction matrix; S402-4: Generate the multi-source cold rolling error collaborative compensation amount by performing actuator physical limit processing on the cold rolling error dynamic compensation distribution correction matrix.

10. A multi-source error collaborative compensation system based on on-line detection of cold-rolled thin plates, the system is applied to the multi-source error collaborative compensation method as described in any one of claims 1-9, characterized in that, It includes a depth-time-space error coupling module, a cold rolling error dynamic weight distribution module, a collaborative compensation coefficient calculation module, and a cold rolling error collaborative compensation module; The depth-time-space error coupling module is used to collect multi-dimensional data of cold rolled thin plates in real time and generate a cold rolling error propagation chain relationship map through a depth-time-space error coupling network; The cold rolling error dynamic weight distribution module is used to establish a cold rolling error dynamic weight distribution mechanism according to the cold rolling error propagation chain relationship map; The collaborative compensation coefficient calculation module is used to calculate the rolling multi-dimensional collaborative compensation coefficient through the cold rolling error dynamic weight distribution mechanism; The cold rolling error collaborative compensation module is used to output the multi-source cold rolling error collaborative compensation amount according to the rolling multi-dimensional collaborative compensation coefficient.

Images