Industrial data lightweight processing method and system based on cloud edge collaboration

Abstract

The invention provides an industrial data lightweight processing method and system based on cloud edge collaboration, and the method comprises the steps: S1, obtaining the multi-modal features of to-be-processed industrial data; s2, based on a scheme decision knowledge base, according to the multi-modal features, a lightweight processing scheme is decided; and S3, based on a cloud edge cooperation technology, according to a lightweight processing scheme, carrying out lightweight processing on the industrial data. According to the method, the lightweight processing scheme is decided based on the scheme decision knowledge base and according to the multi-modal features of the industrial data to be processed, and the industrial data is subjected to lightweight processing based on the cloud edge cooperation technology and according to the lightweight processing scheme. The limitation that the system cannot carry out lightweight processing on the industrial data which is not suitable for the lightweight processing rule due to manual setting of the lightweight processing rule is overcome, the labor cost is reduced, and the lightweight processing efficiency and stability of the industrial data are improved to a great extent.

Description

Technical Field

[0001] The present invention relates to the field of computer data processing technology, and in particular to a method and system for lightweight processing of industrial data based on cloud-edge collaboration. Background Art

[0002] The industrial data generated in industrial environments are usually large and diverse, including sensor data, equipment operating status, etc., which need to be lightweight processed; in industrial production environments, the efficiency and stability of lightweight processing of industrial data directly affect the smoothness and cost-effectiveness of the production process.

[0003] However, when lightweight processing is performed on industrial data, technicians usually manually set lightweight processing rules and the system executes them. However, the system cannot perform lightweight processing on industrial data that does not apply to lightweight processing rules, which has great limitations and cannot guarantee the efficiency and stability of lightweight processing; secondly, the technicians manually set lightweight processing rules, which has a high labor cost.

[0004] Therefore, a solution is urgently needed. Summary of the invention

[0005] The present invention provides a method and system for lightweight processing of industrial data based on cloud-edge collaboration. Based on a solution decision knowledge base, a lightweight processing solution is decided according to the multimodal characteristics of the industrial data to be processed. Based on cloud-edge collaboration technology, the industrial data is lightweight processed according to the lightweight processing solution. This overcomes the limitation that manually set lightweight processing rules make the system unable to perform lightweight processing on industrial data to which the lightweight processing rules are not applicable, reduces labor costs, and greatly improves the efficiency and stability of lightweight processing of industrial data.

[0006] The present invention provides an industrial data lightweight processing method based on cloud-edge collaboration, including:

[0007] Step S1: obtaining multimodal features of industrial data to be processed;

[0008] Step S2: Based on the solution decision knowledge base and multimodal features, a lightweight processing solution is decided;

[0009] Step S3: Based on cloud-edge collaborative technology and according to the lightweight processing solution, perform lightweight processing on industrial data.

[0010] Preferably, the step S2: based on the solution decision knowledge base and according to the multimodal features, deciding on a lightweight processing solution includes:

[0011] Preprocess the multimodal features to obtain preprocessed features;

[0012] Based on the preprocessing features, construct the knowledge applicability conditions;

[0013] Based on the knowledge applicability conditions, target knowledge is selected from the solution decision knowledge base;

[0014] Determine lightweight processing solutions based on target knowledge;

[0015] The preprocessing of the multimodal features to obtain preprocessed features includes:

[0016] When there is a first sub-feature in the multimodal feature that matches the first trigger feature, supplementarily screening out a second sub-feature from the multimodal feature based on the supplementary screening condition corresponding to the first trigger feature;

[0017] The first sub-feature and the second sub-feature are used as preprocessing features;

[0018] and / or,

[0019] Constructing a first target feature set; the first target feature set includes a plurality of third sub-features in the multimodal feature, and the feature type distribution of each third sub-feature in the first target feature set is consistent with the feature type distribution of other preprocessing features based on which other knowledge applicable conditions were constructed in history;

[0020] Using the third sub-feature in the first target feature set as a preprocessing feature;

[0021] and / or,

[0022] Constructing a second target feature set; the second target feature set includes a plurality of fourth sub-features in the multimodal feature set, and each of the fourth sub-features in the second target feature set has a feature association relationship between each other;

[0023] The fourth sub-feature in the second target feature set is used as a preprocessing feature.

[0024] Preferably, the lightweight processing method for industrial data based on cloud-edge collaboration also includes:

[0025] Obtain information on the processing of lightweight industrial data;

[0026] Adaptively adjust the lightweight processing solution based on the processing situation information;

[0027] Based on cloud-edge collaborative technology and according to the adjusted lightweight processing solution, the relay carries out lightweight processing of industrial data.

[0028] Preferably, the adaptive adjustment of the lightweight processing scheme based on the processing situation information includes:

[0029] Based on the standard evaluation system, the processing information is adjusted and evaluated to obtain multiple adjustment requirements and demand levels; the adjustment requirements and demand levels correspond one to one;

[0030] Sort the adjustment requirements from large to small according to the degree of need to obtain the adjustment requirement sequence;

[0031] Traverse the adjustment requirements in the adjustment requirement sequence in sequence;

[0032] When traversing to the i-th adjustment requirement in the adjustment requirement sequence, a requirement screening condition is generated based on the i-th adjustment requirement;

[0033] Based on the demand screening conditions, select the target adjustment demand from the adjustment demand sequence;

[0034] When continuing to traverse the adjustment requirements in the adjustment requirements, the target adjustment requirements are no longer traversed;

[0035] After traversing the adjustment requirements in the adjustment requirement sequence, the target adjustment requirement is removed from the adjustment requirement sequence;

[0036] Perform feature description processing on the adjustment demand sequence after removing the target adjustment demand to obtain a feature description vector;

[0037] Determine an adjustment scheme corresponding to the feature description vector from an adjustment scheme library;

[0038] Based on the adjustment plan, the lightweight processing plan is adaptively adjusted;

[0039] The requirements screening conditions include:

[0040] j>i; j is the order number of the target adjustment demand in the adjustment demand sequence;

[0041] and,

[0042] There is a demand association relationship between the target adjustment demand and the i-th adjustment demand.

[0043] Preferably, the lightweight processing method for industrial data based on cloud-edge collaboration also includes:

[0044] Based on the processing situation information, a situation visualization model is constructed;

[0045] Output visualization model.

[0046] The present invention provides an industrial data lightweight processing system based on cloud-edge collaboration, including:

[0047] An acquisition module, used for acquiring multimodal features of industrial data to be processed;

[0048] The decision module is used to decide on lightweight processing solutions based on the solution decision knowledge base and multimodal features;

[0049] The processing module is used to perform lightweight processing on industrial data based on cloud-edge collaborative technology and according to lightweight processing solutions.

[0050] Preferably, the decision module decides on a lightweight processing solution based on a solution decision knowledge base and multimodal features, including:

[0051] Preprocess the multimodal features to obtain preprocessed features;

[0052] Based on the preprocessing features, construct the knowledge applicability conditions;

[0053] Based on the knowledge applicability conditions, target knowledge is selected from the solution decision knowledge base;

[0054] Determine lightweight processing solutions based on target knowledge;

[0055] The preprocessing of the multimodal features to obtain preprocessed features includes:

[0056] When there is a first sub-feature in the multimodal feature that matches the first trigger feature, supplementarily screening out a second sub-feature from the multimodal feature based on the supplementary screening condition corresponding to the first trigger feature;

[0057] The first sub-feature and the second sub-feature are used as preprocessing features;

[0058] and / or,

[0059] Constructing a first target feature set; the first target feature set includes a plurality of third sub-features in the multimodal feature, and the feature type distribution of each third sub-feature in the first target feature set is consistent with the feature type distribution of other preprocessing features based on which other knowledge applicable conditions were constructed in history;

[0060] Using the third sub-feature in the first target feature set as a preprocessing feature;

[0061] and / or,

[0062] Constructing a second target feature set; the second target feature set includes a plurality of fourth sub-features in the multimodal feature set, and each of the fourth sub-features in the second target feature set has a feature association relationship between each other;

[0063] The fourth sub-feature in the second target feature set is used as a preprocessing feature.

[0064] Preferably, the industrial data lightweight processing system based on cloud-edge collaboration also includes:

[0065] Adjustment modules to include:

[0066] Obtain information on the processing of lightweight industrial data;

[0067] Adaptively adjust the lightweight processing solution based on the processing situation information;

[0068] Based on cloud-edge collaborative technology and according to the adjusted lightweight processing solution, the relay carries out lightweight processing of industrial data.

[0069] Preferably, the adjustment module adaptively adjusts the lightweight processing scheme based on the processing situation information, including:

[0070] Based on the standard evaluation system, the processing information is adjusted and evaluated to obtain multiple adjustment requirements and demand levels; the adjustment requirements and demand levels correspond one to one;

[0071] Sort the adjustment requirements from large to small according to the degree of need to obtain the adjustment requirement sequence;

[0072] Traverse the adjustment requirements in the adjustment requirement sequence in sequence;

[0073] When traversing to the i-th adjustment requirement in the adjustment requirement sequence, a requirement screening condition is generated based on the i-th adjustment requirement;

[0074] Based on the demand screening conditions, select the target adjustment demand from the adjustment demand sequence;

[0075] When continuing to traverse the adjustment requirements in the adjustment requirements, the target adjustment requirements are no longer traversed;

[0076] After traversing the adjustment requirements in the adjustment requirement sequence, the target adjustment requirement is removed from the adjustment requirement sequence;

[0077] Perform feature description processing on the adjustment demand sequence after removing the target adjustment demand to obtain a feature description vector;

[0078] Determine an adjustment scheme corresponding to the feature description vector from an adjustment scheme library;

[0079] Based on the adjustment plan, the lightweight processing plan is adaptively adjusted;

[0080] The requirements screening conditions include:

[0081] j>i; j is the order number of the target adjustment demand in the adjustment demand sequence;

[0082] and,

[0083] There is a demand association relationship between the target adjustment demand and the i-th adjustment demand.

[0084] Preferably, the industrial data lightweight processing system based on cloud-edge collaboration also includes:

[0085] Visualization modules for:

[0086] Based on the processing situation information, a situation visualization model is constructed;

[0087] Output visualization model.

[0088] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0089] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0091] Figure 1 It is a flow chart of a lightweight processing method for industrial data based on cloud-edge collaboration in an embodiment of the present invention;

[0092] Figure 2 This is a schematic diagram of an industrial data lightweight processing system based on cloud-edge collaboration in an embodiment of the present invention. DETAILED DESCRIPTION

[0093] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0094] The present invention provides a lightweight processing method for industrial data based on cloud-edge collaboration, such as Figure 1 As shown, including:

[0095] Step S1: obtaining multimodal features of industrial data to be processed;

[0096] Step S2: Based on the solution decision knowledge base and multimodal features, a lightweight processing solution is decided;

[0097] Step S3: Based on cloud-edge collaborative technology and according to the lightweight processing solution, perform lightweight processing on industrial data.

[0098] The industrial data to be processed refers to the industrial data to be lightweight processed; the multimodal characteristics include: the data type, data volume, data source, and data purpose of the industrial data; there is a large amount of solution decision knowledge in the solution decision knowledge base, and the solution decision knowledge can be used to decide on a lightweight processing solution based on the multimodal characteristics; the solution decision knowledge can be empirical information for deciding on a lightweight processing solution, etc.; the lightweight processing solution is a solution for lightweight processing of industrial data based on cloud-edge collaborative technology; cloud-edge collaborative technology refers to the technology of using cloud computing and edge computing to collaboratively process data, which belongs to the scope of existing technology and will not be elaborated on.

[0099] This application is based on a solution decision knowledge base and decides on a lightweight processing solution according to the multimodal characteristics of the industrial data to be processed. Based on cloud-edge collaborative technology, the industrial data is lightweight processed according to the lightweight processing solution. This overcomes the limitation that manually set lightweight processing rules make the system unable to perform lightweight processing on industrial data to which the lightweight processing rules are not applicable, reduces labor costs, and greatly improves the efficiency and stability of lightweight processing of industrial data.

[0100] In one embodiment, the step S2: based on the solution decision knowledge base and according to the multimodal features, deciding on a lightweight processing solution includes:

[0101] Preprocess the multimodal features to obtain preprocessed features;

[0102] Based on the preprocessing features, construct the knowledge applicability conditions;

[0103] Based on the knowledge applicability conditions, target knowledge is selected from the solution decision knowledge base;

[0104] Determine lightweight processing solutions based on target knowledge;

[0105] The preprocessing of the multimodal features to obtain preprocessed features includes:

[0106] When there is a first sub-feature in the multimodal feature that matches the first trigger feature, supplementarily screening out a second sub-feature from the multimodal feature based on the supplementary screening condition corresponding to the first trigger feature;

[0107] The first sub-feature and the second sub-feature are used as preprocessing features;

[0108] and / or,

[0109] Constructing a first target feature set; the first target feature set includes a plurality of third sub-features in the multimodal feature, and the feature type distribution of each third sub-feature in the first target feature set is consistent with the feature type distribution of other preprocessing features based on which other knowledge applicable conditions were constructed in history;

[0110] Using the third sub-feature in the first target feature set as a preprocessing feature;

[0111] and / or,

[0112] Constructing a second target feature set; the second target feature set includes a plurality of fourth sub-features in the multimodal feature set, and each of the fourth sub-features in the second target feature set has a feature association relationship between each other;

[0113] The fourth sub-feature in the second target feature set is used as a preprocessing feature.

[0114] The preprocessing feature represents the situation of industrialized data that needs to be lightweight processed; the knowledge applicable condition is used to filter out the solution decision knowledge applicable to such situations, that is, the target knowledge; based on the target knowledge, determine the lightweight processing solution; for example: the preprocessing feature is that the data type of industrialized data is product quality inspection data, and the constructed knowledge applicable condition is that the solution decision knowledge must be applicable to the lightweight processing of product quality inspection data. There are three ways to preprocess multimodal features; in the first way, the first trigger feature represents the existence of other second sub-features that can be combined with the first sub-feature to represent the situation of industrialized data that needs to be lightweight processed. For example: the first sub-feature that matches the first trigger feature is that the data type of industrialized data is product quality inspection data, and the second sub-feature is that the data use of product quality inspection data is to report to the superior management department. Therefore, the first sub-feature and the second sub-feature can be combined to represent the situation of industrialized data that needs to be lightweight processed, and the supplementary screening condition is the condition for filtering out the second sub-feature; in the second way, the feature type distribution of the third sub-feature refers to the feature type of each third sub-feature; historically, when constructing other knowledge applicable conditions, it will also be based on other preprocessing features; The feature type distribution of other preprocessing features refers to the feature types of the other preprocessing features; when the feature type distribution of each third sub-feature in the first target feature set is consistent with the feature type distribution of other preprocessing features on which other knowledge applicability conditions have been historically constructed, it means that the third sub-feature in the first target feature set may also be used to construct knowledge applicability conditions, and the third sub-feature in the first target feature set is used as the preprocessing feature; in the third method, the second target feature set is constructed, and when each fourth sub-feature in the second target feature set has a feature association relationship between each other, it means that the fourth features in the second target feature set can be used together to determine the knowledge applicability conditions, for example: the feature association relationship is to jointly represent the main purpose of industrial data, etc., and technical personnel can set the feature association relationship in advance according to actual needs.

[0115] Due to the huge amount of industrial data, the multimodal features of industrial data will also be numerous and complex, which may make the steps of constructing knowledge applicable conditions more cumbersome, reducing the efficiency of subsequent lightweight processing of industrial data. The embodiment of the present invention can solve this problem. Before using multimodal features to construct knowledge applicable conditions, the multimodal features are preprocessed, and the preprocessed multimodal features are used to construct knowledge applicable conditions; in addition, when preprocessing multimodal features, three preprocessing methods are introduced, which greatly improves the rationality, comprehensiveness and adaptability of preprocessing multimodal data.

[0116] In one embodiment, the method for lightweight processing of industrial data based on cloud-edge collaboration further includes:

[0117] Obtain information on the processing of lightweight industrial data;

[0118] Adaptively adjust the lightweight processing solution based on the processing situation information;

[0119] Based on cloud-edge collaborative technology and according to the adjusted lightweight processing solution, the relay carries out lightweight processing of industrial data.

[0120] Processing status information includes: processing progress, processing duration, estimated remaining processing time, etc. Based on the processing status information, the lightweight processing solution is adaptively adjusted; based on cloud-edge collaborative technology, the lightweight processing of industrial data is relayed according to the adjusted lightweight processing solution. Always ensure that the lightweight processing of industrial data is carried out with the best lightweight processing solution, which greatly improves the efficiency of lightweight processing of industrial data.

[0121] In one embodiment, the adaptive adjustment of the lightweight processing scheme based on the processing situation information includes:

[0122] Based on the standard evaluation system, the processing information is adjusted and evaluated to obtain multiple adjustment requirements and demand degrees; the adjustment requirements and demand degrees correspond to each other one by one; the standard evaluation system is a system for adjusting and evaluating the processing information, and the adjustment requirements and corresponding demand degrees can be determined through evaluation; for example: the standard evaluation system is to determine whether the processing progress in the processing information is too slow, and if so, the output adjustment requirement is to speed up the processing progress, and the output corresponding demand degree is 8; the larger the demand degree, the higher the priority of the adjustment requirement to be considered;

[0123] Sort the adjustment requirements from large to small according to the degree of need to obtain the adjustment requirement sequence;

[0124] Traverse the adjustment requirements in the adjustment requirement sequence in sequence;

[0125] When traversing to the i-th adjustment requirement in the adjustment requirement sequence, a requirement screening condition is generated based on the i-th adjustment requirement;

[0126] Based on the demand screening conditions, select the target adjustment demand from the adjustment demand sequence;

[0127] When continuing to traverse the adjustment requirements in the adjustment requirements, the target adjustment requirements are no longer traversed;

[0128] After traversing the adjustment requirements in the adjustment requirement sequence, the target adjustment requirement is removed from the adjustment requirement sequence;

[0129] Perform feature description processing on the adjustment demand sequence after removing the target adjustment demand to obtain a feature description vector;

[0130] Determine an adjustment scheme corresponding to the feature description vector from an adjustment scheme library; the adjustment scheme library contains adjustment schemes corresponding to different feature description vectors; the technicians can set them in advance;

[0131] Based on the adjustment plan, the lightweight processing plan is adaptively adjusted;

[0132] The requirements screening conditions include:

[0133] j>i; j is the order number of the target adjustment demand in the adjustment demand sequence; the order number is the number in the sequence; setting this demand screening condition can make the screened target adjustment demand come after the i-th adjustment demand;

[0134] and,

[0135] There is a demand association relationship between the target adjustment demand and the i-th adjustment demand. The demand association relationship refers to belonging to the same demand type, etc.; when belonging to the same demand type, the demand degree of the i-th adjustment demand is greater than the target adjustment demand, only the i-th adjustment demand is retained, and the target adjustment demand can be eliminated.

[0136] The embodiment of the present invention determines adjustment requirements when adaptively adjusting the lightweight processing solution based on the processing situation information, and performs elimination screening on the adjustment requirements, thereby improving the efficiency of adaptively adjusting the lightweight processing solution.

[0137] In one embodiment, the method for lightweight processing of industrial data based on cloud-edge collaboration further includes:

[0138] Based on the processing situation information, a situation visualization model is constructed;

[0139] Output visualization model.

[0140] It is also possible to build a situation visualization model based on the processing status information and output the situation visualization model to help managers visualize the lightweight processing of industrial data.

[0141] The present invention provides an industrial data lightweight processing system based on cloud-edge collaboration, such as Figure 2 As shown, including:

[0142] An acquisition module 1 is used to acquire multimodal features of industrial data to be processed;

[0143] Decision module 2, used to decide lightweight processing solutions based on solution decision knowledge base and multimodal features;

[0144] Processing module 3 is used to perform lightweight processing on industrial data based on cloud-edge collaborative technology and a lightweight processing solution.

[0145] The decision module 2 decides a lightweight processing solution based on the solution decision knowledge base and the multimodal features, including:

[0146] Preprocess the multimodal features to obtain preprocessed features;

[0147] Based on the preprocessing features, construct the knowledge applicability conditions;

[0148] Based on the knowledge applicability conditions, target knowledge is selected from the solution decision knowledge base;

[0149] Determine lightweight processing solutions based on target knowledge;

[0150] The preprocessing of the multimodal features to obtain preprocessed features includes:

[0151] When there is a first sub-feature in the multimodal feature that matches the first trigger feature, supplementarily screening out a second sub-feature from the multimodal feature based on the supplementary screening condition corresponding to the first trigger feature;

[0152] The first sub-feature and the second sub-feature are used as preprocessing features;

[0153] and / or,

[0154] Constructing a first target feature set; the first target feature set includes a plurality of third sub-features in the multimodal feature, and the feature type distribution of each third sub-feature in the first target feature set is consistent with the feature type distribution of other preprocessing features based on which other knowledge applicable conditions were constructed in history;

[0155] Using the third sub-feature in the first target feature set as a preprocessing feature;

[0156] and / or,

[0157] Constructing a second target feature set; the second target feature set includes a plurality of fourth sub-features in the multimodal feature set, and each of the fourth sub-features in the second target feature set has a feature association relationship between each other;

[0158] The fourth sub-feature in the second target feature set is used as a preprocessing feature.

[0159] The lightweight industrial data processing system based on cloud-edge collaboration also includes:

[0160] Adjustment modules to include:

[0161] Obtain information on the processing of lightweight industrial data;

[0162] Adaptively adjust the lightweight processing solution based on the processing situation information;

[0163] Based on cloud-edge collaborative technology and according to the adjusted lightweight processing solution, the relay carries out lightweight processing of industrial data.

[0164] The adjustment module adaptively adjusts the lightweight processing solution based on the processing situation information, including:

[0165] Based on the standard evaluation system, the processing information is adjusted and evaluated to obtain multiple adjustment requirements and demand levels; the adjustment requirements and demand levels correspond one to one;

[0166] Sort the adjustment requirements from large to small according to the degree of need to obtain the adjustment requirement sequence;

[0167] Traverse the adjustment requirements in the adjustment requirement sequence in sequence;

[0168] When traversing to the i-th adjustment requirement in the adjustment requirement sequence, a requirement screening condition is generated based on the i-th adjustment requirement;

[0169] Based on the demand screening conditions, select the target adjustment demand from the adjustment demand sequence;

[0170] When continuing to traverse the adjustment requirements in the adjustment requirements, the target adjustment requirements are no longer traversed;

[0171] After traversing the adjustment requirements in the adjustment requirement sequence, the target adjustment requirement is removed from the adjustment requirement sequence;

[0172] Perform feature description processing on the adjustment demand sequence after removing the target adjustment demand to obtain a feature description vector;

[0173] Determine an adjustment scheme corresponding to the feature description vector from an adjustment scheme library;

[0174] Based on the adjustment plan, the lightweight processing plan is adaptively adjusted;

[0175] The requirements screening conditions include:

[0176] j>i; j is the order number of the target adjustment demand in the adjustment demand sequence;

[0177] and,

[0178] There is a demand association relationship between the target adjustment demand and the i-th adjustment demand.

[0179] The lightweight industrial data processing system based on cloud-edge collaboration also includes:

[0180] Visualization modules for:

[0181] Based on the processing situation information, a situation visualization model is constructed;

[0182] Output visualization model.

[0183] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A lightweight processing method for industrial data based on cloud-edge collaboration, characterized in that: include: Step S1: obtaining multimodal features of industrial data to be processed; Step S2: Based on the solution decision knowledge base and multimodal features, a lightweight processing solution is decided; Step S3: Based on cloud-edge collaborative technology and according to the lightweight processing solution, perform lightweight processing on industrial data.

2. The method for lightweight processing of industrial data based on cloud-edge collaboration according to claim 1, characterized in that: The step S2: based on the solution decision knowledge base and according to the multimodal features, a lightweight processing solution is decided, including: Preprocess the multimodal features to obtain preprocessed features; Based on the preprocessing features, construct the knowledge applicability conditions; Based on the knowledge applicability conditions, target knowledge is selected from the solution decision knowledge base; Determine lightweight processing solutions based on target knowledge; The preprocessing of the multimodal features to obtain preprocessed features includes: When there is a first sub-feature in the multimodal feature that matches the first trigger feature, supplementarily screening out a second sub-feature from the multimodal feature based on the supplementary screening condition corresponding to the first trigger feature; The first sub-feature and the second sub-feature are used as preprocessing features; and / or, Constructing a first target feature set; the first target feature set includes a plurality of third sub-features in the multimodal feature, and the feature type distribution of each third sub-feature in the first target feature set is consistent with the feature type distribution of other preprocessing features based on which other knowledge applicable conditions were constructed in history; Using the third sub-feature in the first target feature set as a preprocessing feature; and / or, Constructing a second target feature set; the second target feature set includes a plurality of fourth sub-features in the multimodal feature set, and each of the fourth sub-features in the second target feature set has a feature association relationship between each other; The fourth sub-feature in the second target feature set is used as a preprocessing feature.

3. The method for lightweight processing of industrial data based on cloud-edge collaboration according to claim 1, characterized in that: Also includes: Obtain information on the processing of lightweight industrial data; Adaptively adjust the lightweight processing solution based on the processing situation information; Based on cloud-edge collaborative technology and according to the adjusted lightweight processing solution, the relay carries out lightweight processing of industrial data.

4. The method for lightweight processing of industrial data based on cloud-edge collaboration according to claim 3 is characterized in that: The adaptive adjustment of the lightweight processing scheme based on the processing situation information includes: Based on the standard evaluation system, the processing information is adjusted and evaluated to obtain multiple adjustment requirements and demand levels; the adjustment requirements and demand levels correspond one to one; Sort the adjustment requirements from large to small according to the degree of need to obtain the adjustment requirement sequence; Traverse the adjustment requirements in the adjustment requirement sequence in sequence; When traversing to the i-th adjustment requirement in the adjustment requirement sequence, a requirement screening condition is generated based on the i-th adjustment requirement; Based on the demand screening conditions, select the target adjustment demand from the adjustment demand sequence; When continuing to traverse the adjustment requirements in the adjustment requirements, the target adjustment requirements are no longer traversed; After traversing the adjustment requirements in the adjustment requirement sequence, the target adjustment requirement is removed from the adjustment requirement sequence; Perform feature description processing on the adjustment demand sequence after removing the target adjustment demand to obtain a feature description vector; Determine an adjustment scheme corresponding to the feature description vector from an adjustment scheme library; Based on the adjustment plan, the lightweight processing plan is adaptively adjusted; The requirements screening conditions include: j>i; j is the order number of the target adjustment demand in the adjustment demand sequence; and, There is a demand association relationship between the target adjustment demand and the i-th adjustment demand.

5. The method for lightweight processing of industrial data based on cloud-edge collaboration according to claim 3 is characterized in that: Also includes: Based on the processing situation information, a situation visualization model is constructed; Output visualization model.

6. The lightweight industrial data processing system based on cloud-edge collaboration is characterized by: include: An acquisition module, used for acquiring multimodal features of industrial data to be processed; The decision module is used to decide on lightweight processing solutions based on the solution decision knowledge base and multimodal features; The processing module is used to perform lightweight processing on industrial data based on cloud-edge collaborative technology and according to lightweight processing solutions.

7. The industrial data lightweight processing system based on cloud-edge collaboration according to claim 6 is characterized in that: The decision module decides on a lightweight processing solution based on a solution decision knowledge base and multimodal features, including: Preprocess the multimodal features to obtain preprocessed features; Based on the preprocessing features, construct the knowledge applicability conditions; Based on the knowledge applicability conditions, target knowledge is selected from the solution decision knowledge base; Determine lightweight processing solutions based on target knowledge; The preprocessing of the multimodal features to obtain preprocessed features includes: When there is a first sub-feature in the multimodal feature that matches the first trigger feature, supplementarily screening out a second sub-feature from the multimodal feature based on the supplementary screening condition corresponding to the first trigger feature; The first sub-feature and the second sub-feature are used as preprocessing features; and / or, Constructing a first target feature set; the first target feature set includes a plurality of third sub-features in the multimodal feature, and the feature type distribution of each third sub-feature in the first target feature set is consistent with the feature type distribution of other preprocessing features based on which other knowledge applicable conditions were constructed in history; Using the third sub-feature in the first target feature set as a preprocessing feature; and / or, Constructing a second target feature set; the second target feature set includes a plurality of fourth sub-features in the multimodal feature set, and each of the fourth sub-features in the second target feature set has a feature association relationship between each other; The fourth sub-feature in the second target feature set is used as a preprocessing feature.

8. The industrial data lightweight processing system based on cloud-edge collaboration according to claim 6 is characterized in that: Also includes: Adjustment modules to include: Obtain information on the processing of lightweight industrial data; Adaptively adjust the lightweight processing solution based on the processing situation information; Based on cloud-edge collaborative technology and according to the adjusted lightweight processing solution, the relay carries out lightweight processing of industrial data.

9. The industrial data lightweight processing system based on cloud-edge collaboration as claimed in claim 8, characterized in that: The adjustment module adaptively adjusts the lightweight processing solution based on the processing situation information, including: Based on the standard evaluation system, the processing information is adjusted and evaluated to obtain multiple adjustment requirements and demand levels; the adjustment requirements and demand levels correspond one to one; Sort the adjustment requirements from large to small according to the degree of need to obtain the adjustment requirement sequence; Traverse the adjustment requirements in the adjustment requirement sequence in sequence; When traversing to the i-th adjustment requirement in the adjustment requirement sequence, a requirement screening condition is generated based on the i-th adjustment requirement; Based on the demand screening conditions, select the target adjustment demand from the adjustment demand sequence; When continuing to traverse the adjustment requirements in the adjustment requirements, the target adjustment requirements are no longer traversed; After traversing the adjustment requirements in the adjustment requirement sequence, the target adjustment requirement is removed from the adjustment requirement sequence; Perform feature description processing on the adjustment demand sequence after removing the target adjustment demand to obtain a feature description vector; Determine an adjustment scheme corresponding to the feature description vector from an adjustment scheme library; Based on the adjustment plan, the lightweight processing plan is adaptively adjusted; The requirements screening conditions include: j>i; j is the order number of the target adjustment demand in the adjustment demand sequence; and, There is a demand association relationship between the target adjustment demand and the i-th adjustment demand.

10. The industrial data lightweight processing system based on cloud-edge collaboration according to claim 8, characterized in that: Also includes: Visualization modules for: Based on the processing situation information, a situation visualization model is constructed; Output visualization model.

Images