Method and device for determining measurement deviation of belt weigher in cut tobacco making process
By calculating the cumulative weight approximation value of the belt scale in the wire making process using multiple prediction models, the problem of difficult to evaluate the measurement deviation of the belt scale is solved, and the accuracy of loss point positioning in the wire making process is improved.
Patent Information
- Application Number
- CN202411532759.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-30
- Publication Date
- 2025-02-18
AI Technical Summary
In the wire making process, the measurement deviation of the belt scale is difficult to accurately evaluate, resulting in the inability to find the loss points in time and accurately, causing great waste and losses to the factory.
The predicted value of the accumulated weight of the belt scale is obtained by multiple different prediction models, and the approximation value of the actual accumulated weight is calculated based on the multiple prediction values, thereby evaluating the measurement deviation of the belt scale.
It improves the accuracy of deviation evaluation of belt scales, which is conducive to accurately determining the loss points in the wire making process and reduces factory waste and losses.
Smart Images
Figure CN119469357A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of silk production, and more specifically, to a method and device for determining measurement deviation of a belt scale in a silk production process. Background Art
[0002] In the process of making cigarettes, from the original cigarette box leaving the warehouse to the finished tobacco entering the cabinet, many key processes are required in the middle. The entire processing process is cumbersome, the process route is long, and there are many loss points. Lean production and cost reduction and efficiency improvement are also effective ways and urgent needs for factories to improve their competitiveness. Among them, the wire yield rate is a key quality indicator of the wire making process. Due to the lack of precise control of each key process, when the wire yield rate drops, it is impossible to find the loss point in time and accurately, causing great waste and loss to the factory.
[0003] Online real-time monitoring of the cumulative weight deviation of tobacco leaves on the silk production line is an important part of realizing the intelligent silk production line. Real-time monitoring requires comparing the predicted value and measured value of the cumulative weight of the belt scale, but both have a certain deviation from the actual cumulative weight. Since the actual cumulative weight can never be accurately obtained in practice, when there is a deviation between the predicted value and the measured value, it is impossible to determine which value is closer to the actual cumulative weight. If the measured value of the belt scale has a large deviation, how to calculate the deviation data? There is currently no relevant technology in the industry. Summary of the invention
[0004] The present application provides a method and device for determining the measurement deviation of a belt scale in a silk-making process, which obtains a predicted value of the cumulative weight of the belt scale through multiple different prediction models, and calculates an approximate value of the actual cumulative weight based on the multiple predicted values, thereby using the approximate value as the actual cumulative weight to evaluate the measurement deviation, thereby improving the accuracy of the deviation evaluation of the belt scale and facilitating the accurate determination of the loss points in the silk-making process.
[0005] The present application provides a method for determining the measurement deviation of a belt scale in a silk making process, comprising:
[0006] Receive process parameters related to the accumulated weight of the belt scale collected during silk production;
[0007] Using the process parameters to obtain multiple first predicted values of the cumulative weight of the target belt scale in the silk making process, the multiple first predicted values are obtained based on different prediction models;
[0008] Determine an approximate value of the actual cumulative weight of the target belt scale based on the plurality of first predicted values;
[0009] The measured deviation of the accumulated weight of the target belt scale is evaluated based on the approximation value.
[0010] Preferably, the multiple prediction models are different prediction models established based on different process position relationships between the silk making process flow and the target belt scale.
[0011] Preferably, determining an approximate value of the actual cumulative weight of the target belt scale based on the multiple first predicted values specifically includes:
[0012] Removing the maximum value and the minimum value from the plurality of first prediction values to obtain a plurality of second prediction values;
[0013] Recalculate the second weight corresponding to each second prediction value based on the first weight corresponding to each first prediction value;
[0014] An approximation value is calculated based on all second predicted values and their second weights.
[0015] Preferably, the measurement deviation of the cumulative weight of the target belt scale is evaluated based on the approximation value, specifically including:
[0016] Calculate the approximate value of the target belt scale for each batch;
[0017] If the approximate values of a preset number of batches are all greater than the measured values of the target belt scale or are all less than the measured values of the target belt scale, it is determined that the target belt scale has a weighing deviation of being too heavy or too light.
[0018] Preferably, the method of evaluating the measurement deviation of the cumulative weight of the target belt scale based on the approximation value further includes:
[0019] If the absolute value of the difference between the approximation value and the measured value of a single batch of target belt scales is greater than a threshold, the difference between the measured value and the approximation value is taken as the measurement deviation of the batch of target belt scales.
[0020] The present application also provides a device for determining measurement deviation of a belt scale in a silk-making process, comprising a receiving module, a prediction module, an approximate value determination module, and an evaluation module;
[0021] The receiving module is used to receive the process parameters related to the accumulated weight of the belt scale collected during the silk production;
[0022] The prediction module is used to obtain multiple first prediction values of the cumulative weight of the target belt scale in the silk making process by using the process parameters, and the multiple first prediction values are obtained based on different prediction models;
[0023] The approximation value determination module is used to determine the approximation value of the actual cumulative weight of the target belt scale based on the multiple first predicted values;
[0024] The evaluation module is used for evaluating the measurement deviation of the accumulated weight of the target belt scale based on the approximation value.
[0025] Preferably, the multiple prediction models are different prediction models established based on different process position relationships between the silk making process flow and the target belt scale.
[0026] Preferably, the approximation value determination module includes an extreme value removal module, a weight determination module and a first approximation value calculation module;
[0027] The extreme value removal module is used to remove the maximum value and the minimum value from the plurality of first prediction values to obtain a plurality of second prediction values;
[0028] The weight determination module is used to recalculate the second weight corresponding to each second prediction value based on the first weight corresponding to each first prediction value;
[0029] The first approximation value calculation module is used to calculate the approximation value based on all the second prediction values and their second weights.
[0030] Preferably, the evaluation module includes a second approximation value calculation module and a qualitative determination module;
[0031] The second approximation value calculation module is used to calculate the approximation value of each batch of target belt scales;
[0032] The qualitative judgment module is used to determine that the target belt scale has a weighing deviation of being too heavy or too light when the approximate values of a preset number of batches are all greater than the measured value of the target belt scale or are all less than the measured value of the target belt scale.
[0033] Preferably, the evaluation module also includes a quantitative calculation module, which is used to use the difference between the measured value and the approximation value as the measurement deviation of the target belt scale of a single batch when the absolute value of the difference between the approximation value and the measured value of the target belt scale of the batch is greater than a threshold.
[0034] Other features and advantages of the present application will become apparent from the following detailed description of exemplary embodiments of the present application with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the application and, together with the description, serve to explain the principles of the application.
[0036] Figure 1 A flow chart of a method for determining measurement deviation of a belt scale in a silk-making process provided in the present application;
[0037] Figure 2 An example of the silk making process and process parameters provided for this application;
[0038] Figure 3 A schematic diagram of the front-end full-line prediction model provided for this application;
[0039] Figure 4A schematic diagram of the back-end full-line prediction model provided for this application;
[0040] Figure 5 A schematic diagram of the whole-line prediction model of the silk-making process provided in this application;
[0041] Figure 6 A schematic diagram of the front-end single-point prediction model provided for this application;
[0042] Figure 7 A schematic diagram of the backend single-point prediction model provided for this application;
[0043] Figure 8 A schematic diagram of the back-end multi-point prediction model provided for this application;
[0044] Fig. 9 This is a structural diagram of a device for determining measurement deviation of a belt scale in a silk making process provided in this application. DETAILED DESCRIPTION
[0045] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present application.
[0046] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present application, its application, or uses.
[0047] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.
[0048] In all examples shown and discussed herein, any specific values should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.
[0049] The present application provides a method and device for determining the measurement deviation of a belt scale in a silk-making process, which obtains a predicted value of the cumulative weight of the belt scale through multiple different prediction models, and calculates an approximate value of the actual cumulative weight based on the multiple predicted values, thereby using the approximate value as the actual cumulative weight to evaluate the measurement deviation, thereby improving the accuracy of the deviation evaluation of the belt scale and facilitating the accurate determination of the loss points in the silk-making process.
[0050] like Figure 1 As shown, the method for determining the measurement deviation of the belt scale in the silk making process provided by the present application includes:
[0051] S110: receiving process parameters related to the accumulated weight of the belt scale collected during silk production.
[0052] Specifically, first draw a silk making process flow chart, and then mark the process parameters related to the cumulative weight of the belt scale on the flow chart, including each weighing belt scale, material moisture measurement point, temperature measurement point, storage cabinet time, temperature and humidity, etc.
[0053] Figure 2 An example of the silk making process and process parameters is shown. Figure 2 As shown in the figure, there are 4 belt scales in the silk making process. All process parameters before belt scale 1 are defined as X 1N Variable group; all process parameters after belt scale 1 and before belt scale 2 are defined as X 2N Variable group; and so on, define X 3N Variable group, X 4N Variable group and X 5N Variable groups. The variables in each variable group can be deleted or added based on the actual situation on site.
[0054] As an example, for X 1N The variable group includes material moisture measurement points, temperature measurement points, storage time, temperature and humidity in the silk slicing section, etc. For example, X 11 The variable is the cumulative amount of water added to loosen and moisten leaves; X 12 The variable is the cumulative value of the loose leaf scale before moistening; X 13 The variable is the loose leaf outlet temperature; X 14 The variable is the moisture value at the loose leaf outlet; X 15 The variable is the moisture value of the leaf before cutting; X 16 The variable is the storage time; X 17 The variable is the time for changing cabinets; X 18 The variable is the atmospheric temperature; X 19 The variable is atmospheric humidity.
[0055] For X 2N The variable group includes material moisture measurement points, temperature measurement points, storage time, temperature and humidity in the primary feeding section of silk making, etc. For example, X 21 The variable is the cumulative amount of material and water added at one time; X 22 The variable is the accumulated value of the scale before adding material once; X 23 The variable is the outlet temperature of the primary feed; X 24 The variable is the moisture value at the outlet of the primary feed; X 25 The variable is the moisture value at the inlet of the primary feed; X 26 The variable is the storage time; X 27 The variable is the time of changing cabinets from pre-distribution cabinet to distribution cabinet; X 28 The variable is the atmospheric temperature; X 29 The variable is atmospheric humidity.
[0056] X3N Variable group, X 4N Variable group and X 5N The factor selection method of the variable groups is consistent.
[0057] The weighing weights (ie, the measured values of the cumulative weights) of the belt scale 1, the belt scale 2, the belt scale 3 and the belt scale 4 are defined as Y1, Y2, Y3 and Y4 respectively.
[0058] S120: using the process parameters to obtain a plurality of first predicted values of the cumulative weight of the target belt scale in the silk making process, wherein the plurality of first predicted values are obtained based on different prediction models.
[0059] Preferably, the multiple prediction models are different prediction models established based on different process position relationships between the silk making process and the target belt scale, wherein the process position relationship is the position relationship between the target belt scale and each production line section and other belt scales in the silk making process.
[0060] exist Figure 2 In the example, the following is explained by taking belt scale 2 as the target belt scale.
[0061] As an embodiment, the multiple prediction models include a front-end whole-line prediction model G1 of the target belt scale, a back-end whole-line prediction model G3, and a whole-line prediction model G6 of the silk making process.
[0062] like Figure 3 As shown in the figure, the input data of the front-end whole-line prediction model G1 is all the process parameters of the front-end of the belt scale 2, including X 1N Variable group, weighing weight of belt scale 1 Y1, X 2N Variable group: The output data of the front-end whole-line prediction model G1 is the predicted value of the cumulative weight of the belt scale 2.
[0063] like Figure 4 As shown, the input data of the back-end whole-line prediction model G3 is all the process parameters of the back-end of the belt scale 2, including X 3N Variable group, weighing weight of belt scale 3 Y3, X 4N Variable group, belt scale 4 weighing weight Y4, X 5N The output data of the back-end whole-line prediction model G3 is the predicted value of the accumulated weight of the belt scale 2.
[0064] like Figure 5 As shown in the figure, the input data of the whole line prediction model G6 of the silk making process are all the process parameters of the silk making process, including X 1N Variable group, weighing weight of belt scale 1 Y1, X 2N Variable group, X 3N Variable group, weighing weight of belt scale 3 Y3, X 4NVariable group, belt scale 4 weighing weight Y4, X 5N Variable group: The output data of the whole-line prediction model G6 of the silk-making process is the predicted value of the cumulative weight of the belt scale 2.
[0065] Preferably, the multiple prediction models also include single-point prediction models and multi-point prediction models for the front end and the back end of the target belt scale, where one belt scale corresponds to one point.
[0066] exist Figure 2 In the example, there is a belt scale (i.e., belt scale 1) at the front end of belt scale 2, and the front end of belt scale 2 only involves the front end single-point prediction model G2 of belt scale 1-belt scale 2. There are belt scale 3 and belt scale 4 at the back end of belt scale 2, and the back end of belt scale 2 involves the back end single-point prediction model G4 of belt scale 2-belt scale 3 and the back end multi-point prediction model G5 of belt scale 2-belt scale 3-belt scale 4.
[0067] like Figure 6 As shown in the figure, the input data of the front-end single-point prediction model G2 is the process parameters of belt scale 1-belt scale 2, including the weighing weight Y1 and X 2N Variable group. The output data of the front-end single-point prediction model G2 is the predicted value of the cumulative weight of the belt scale 2.
[0068] like Figure 7 As shown, the input data of the back-end single-point prediction model G4 is the process parameters of belt scale 2-belt scale 3, including X 3N Variable group, weighing weight Y3 of belt scale 3. The output data of back-end single-point prediction model G4 is the predicted value of the cumulative weight of belt scale 2.
[0069] like Figure 8 As shown, the input data of the back-end multi-point prediction model G5 are the process parameters of belt scale 2-belt scale 4, including X 3N Variable group, weighing weight of belt scale 3 Y3, X 4N Variable group, weighing weight Y4 of belt scale 4. The output data of back-end multi-point prediction model G5 is the predicted value of the cumulative weight of belt scale 2.
[0070] It should be noted that the above prediction models are all machine learning models. When constructing each prediction model, the most suitable machine learning model algorithm is selected through comparison as the final prediction model. When training the above prediction models, data preprocessing (such as removing outliers, etc.) is required.
[0071] In this step (S120), the corresponding process parameters are input into each prediction model to obtain a first prediction value of the cumulative weight of a target belt scale, which is recorded as Y 21 , Y 22、 Y 23、Y 24、 Y 25、 Y 26 .
[0072] S130: Determine an approximate value of the actual cumulative weight of the target belt scale based on the multiple first predicted values.
[0073] Based on the knowledge of probability and logic, according to the multiple predicted values of the target belt scale (such as Y 21 , Y 22、 Y 23、 Y 24、 Y 25、 Y 26 data set) to obtain an approximate value This value can be close to the actual cumulative weight of the target belt scale, and is more likely to be closer to the actual cumulative weight than a single predicted value. Therefore, the measurement accuracy of the belt scale can be judged by the approximate value.
[0074] It should be noted that, while constructing the above-mentioned multiple prediction models, a corresponding weight value is given to each prediction model according to its prediction effect on the measured value. Figure 2 In the example shown, the weights of the six prediction models are K 21 , K 22 , K 23 , K 24 , K 25 , K 26 . K 21 +K 22 +K 23 +K 24 +K 25 +K 26 =1.0 and weight value 0 <K 2A <0.4,(A=1…6).
[0075] As an embodiment, determining an approximate value of the actual cumulative weight of the target belt scale based on the multiple first predicted values specifically includes:
[0076] S1301: Remove the maximum value Y among the multiple first prediction values 2i (corresponding weight K 2i ) and minimum value Y 2j (corresponding weight K 2j ), obtain multiple second prediction values, the second prediction value is recorded as Y 2k (corresponding weight K 2k ).
[0077] S1302: Recalculate the second weight corresponding to each second prediction value based on the first weight corresponding to each first prediction value
[0078]
[0079] The sum of the weights of all second prediction values is 1.
[0080] S1303: Calculate an approximation value based on all second prediction values and their second weights
[0081]
[0082] The following is an example to further illustrate the calculation of the approximate value:
[0083] According to the prediction effect of prediction models G1-G6 on Y2 value, weight values K are given to prediction models G1-G6 respectively. 21 =0.15, K 22 =0.2, K 23 =0.15, K 24 =0.2, K 25 =0.10, K 26 =0.2, the corresponding predicted values are Y 21 , Y 22 , Y 23 , Y 24 , Y 25 , Y 26 If Y in the data set 21 , Y 22 , Y 23 , Y 24 , Y 25 , Y 26 The data is sorted by value as Y 21 >Y 22 >Y 23 >Y 24 >Y 25 >Y 26 , remove the maximum value Y 21 and minimum value Y 26 , and then readjust the weight values.
[0084]
[0085] Approximation for:
[0086]
[0087] S140: Evaluate the measurement deviation of the cumulative weight of the target belt scale based on the approximation value.
[0088] As an embodiment, the measurement deviation of the cumulative weight of the target belt scale is evaluated based on the approximation value, specifically including:
[0089] S1401: Calculate the approximate value of the target belt scale for each batch
[0090] S1402: If the approximate value of the single batch target belt scale If the absolute value of the difference between the measured value Y2 and the approximate value is greater than the threshold, it is determined that the target belt scale has a weighing deviation in this batch, and the measured value Y2 is compared with the approximate value The difference is taken as the measurement deviation S of the target belt scale of this batch:
[0091]
[0092] S1403: If the approximate value of the preset number of batches is If both are greater than the measured value Y2 of the target belt scale or both are less than the measured value Y2 of the target belt scale, it is determined that the target belt scale has a weighing deviation of being too heavy or too light.
[0093] It should be noted that if it is determined that the target belt scale has a weighing deviation in a certain batch or as a whole, the belt scale needs to be tared or re-calibrated.
[0094] Based on the above, the present application provides a device for determining the measurement deviation of a belt scale in a silk making process. Fig. 9 As shown, the measurement deviation determination device includes a receiving module 910 , a prediction module 920 , an approximation value determination module 930 and an evaluation module 940 .
[0095] The receiving module 910 is used to receive process parameters related to the accumulated weight of the belt scale collected during silk production.
[0096] The prediction module 920 is used to obtain multiple first prediction values of the cumulative weight of the target belt scale in the silk making process flow by using the process parameters, and the multiple first prediction values are obtained based on different prediction models.
[0097] The approximation value determination module 930 is used to determine an approximation value of the actual cumulative weight of the target belt scale based on a plurality of first prediction values.
[0098] The evaluation module 940 is used to evaluate the measurement deviation of the cumulative weight of the target belt scale based on the approximation value.
[0099] Preferably, the multiple prediction models are different prediction models established based on different process position relationships between the silk making process flow and the target belt scale.
[0100] Preferably, the approximation value determination module 930 includes an extreme value removal module 9301 , a weight determination module 9302 and a first approximation value calculation module 9303 .
[0101] The extreme value removal module 9301 is used to remove the maximum value and the minimum value from the multiple first prediction values to obtain multiple second prediction values.
[0102] The weight determination module 9302 is used to recalculate the second weight corresponding to each second prediction value based on the first weight corresponding to each first prediction value.
[0103] The first approximation value calculation module 9303 is used to calculate the approximation value based on all the second prediction values and their second weights.
[0104] Preferably, the evaluation module 940 includes a second approximation value calculation module 9401 and a qualitative determination module 9402 .
[0105] The second approximation value calculation module 9401 is used to calculate the approximation value of each batch of target belt scales.
[0106] The qualitative determination module 9402 is used to determine that the target belt scale has a weighing deviation of being too heavy or too light when the approximate values of a preset number of consecutive batches are all greater than the measured value of the target belt scale or are all less than the measured value of the target belt scale.
[0107] Preferably, the evaluation module 940 also includes a quantitative calculation module 9403, which is used to use the difference between the measured value and the approximation value as the measurement deviation of the target belt scale of a single batch when the absolute value of the difference between the approximation value and the measured value of the target belt scale of the batch is greater than a threshold.
[0108] The beneficial effects of this application are as follows:
[0109] 1. This application considers the weighing accuracy of the belt scale based on the entire silk making process, and uses multiple predicted values and the measured values of the belt scale to judge the measurement deviation and determine the deviation value, thereby solving the practical problem that errors occur between the predicted values and the measured values and the cause of the errors cannot be further analyzed qualitatively or quantitatively.
[0110] 2. This application utilizes the rules of multiple predicted data to derive the approximate value of the actual cumulative weight, thereby improving the weighing accuracy of the belt scale.
[0111] 3. This application uses both qualitative and quantitative methods to determine the accuracy of belt scale weighing, which improves the reliability and practicality of predicting and diagnosing equipment status.
[0112] Although some specific embodiments of the present application have been described in detail by way of example, it should be understood by those skilled in the art that the above examples are only for illustration, not for limiting the scope of the present application. It should be understood by those skilled in the art that the above embodiments may be modified without departing from the scope and spirit of the present application. The scope of the present application is defined by the appended claims.
Claims
1. A method for determining the measurement deviation of a belt scale in a silk-making process, characterized in that: include: Receive process parameters related to the accumulated weight of the belt scale collected during silk production; Using the process parameters to obtain a plurality of first predicted values of the cumulative weight of a target belt scale in a silk making process, wherein the plurality of first predicted values are obtained based on different prediction models; Determine an approximate value of the actual cumulative weight of the target belt scale based on the plurality of first predicted values; A measured deviation of the cumulative weight of the target belt scale is evaluated based on the approximation value.
2. The method for determining the measurement deviation of a belt scale in a silk making process according to claim 1, characterized in that: The multiple prediction models are different prediction models established based on different process position relationships between the silk making process and the target belt scale.
3. The method for determining the measurement deviation of a belt scale in a silk making process according to claim 1, characterized in that: Determining an approximate value of the actual cumulative weight of the target belt scale based on the multiple first predicted values specifically includes: Removing the maximum value and the minimum value from the plurality of first prediction values to obtain a plurality of second prediction values; Recalculate the second weight corresponding to each second prediction value based on the first weight corresponding to each first prediction value; The approximation value is calculated based on all second prediction values and their second weights.
4. The method for determining the measurement deviation of a belt scale in a silk making process according to claim 1, characterized in that: The measurement deviation of the cumulative weight of the target belt scale is evaluated based on the approximation value, specifically comprising: Calculate the approximate value of the target belt scale for each batch; If the approximate values of a preset number of consecutive batches are all greater than the measured value of the target belt scale or are all less than the measured value of the target belt scale, it is determined that the target belt scale has a weighing deviation of being too heavy or too light.
5. The method for determining the measurement deviation of a belt scale in a silk-making process according to claim 4, characterized in that: The method further comprises: evaluating the measurement deviation of the cumulative weight of the target belt scale based on the approximation value; If the absolute value of the difference between the approximation value and the measured value of a single batch of target belt scales is greater than a threshold, the difference between the measured value and the approximation value is used as the measurement deviation of the target belt scale of the batch.
6. A device for determining measurement deviation of a belt scale in a silk-making process, characterized in that: It includes a receiving module, a prediction module, an approximation value determination module and an evaluation module; The receiving module is used to receive process parameters related to the cumulative weight of the belt scale collected during silk production; The prediction module is used to obtain multiple first prediction values of the cumulative weight of the target belt scale in the silk making process using the process parameters, and the multiple first prediction values are obtained based on different prediction models; The approximation value determination module is used to determine the approximation value of the actual cumulative weight of the target belt scale based on the multiple first predicted values; The evaluation module is used to evaluate the measurement deviation of the cumulative weight of the target belt scale based on the approximation value.
7. The device for determining the measurement deviation of a belt scale in a silk-making process according to claim 6, characterized in that: The multiple prediction models are different prediction models established based on different process position relationships between the silk making process and the target belt scale.
8. The device for determining the measurement deviation of a belt scale in a silk-making process according to claim 6, characterized in that: The approximation value determination module includes an extreme value removal module, a weight determination module and a first approximation value calculation module; The extreme value removal module is used to remove the maximum value and the minimum value from the plurality of first predicted values to obtain a plurality of second predicted values; The weight determination module is used to recalculate the second weight corresponding to each second prediction value based on the first weight corresponding to each first prediction value; The first approximation value calculation module is used to calculate the approximation value based on all second prediction values and their second weights.
9. The device for determining the measurement deviation of a belt scale in a silk-making process according to claim 6, characterized in that: The evaluation module includes a second approximation value calculation module and a qualitative determination module; The second approximation value calculation module is used to calculate the approximation value of each batch of target belt scales; The qualitative determination module is used to determine that the target belt scale has a weighing deviation of being too heavy or too light when the approximate values of a preset number of consecutive batches are all greater than the measured value of the target belt scale or are all less than the measured value of the target belt scale.
10. The device for determining the measurement deviation of a belt scale in a silk-making process according to claim 9, characterized in that: The evaluation module also includes a quantitative calculation module, which is used to use the difference between the measured value and the approximation value as the measurement deviation of the target belt scale of the batch when the absolute value of the difference between the approximation value and the measured value of the target belt scale of the batch is greater than a threshold.