Iron and steel enterprise scientific research project cost collection method and system

Through image processing and clustering models, the problem of low cost collection efficiency of scientific research projects in steel enterprises is solved, and more accurate and efficient cost collection is achieved.

CN118917802APending Publication Date: 2024-11-08BAOTOU IRON & STEEL (GROUP) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411015124.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The collection efficiency of scientific research project costs of steel enterprises is low, and it is easy to cause inaccuracy due to human errors.

Method used

Image processing technology is used to convert the cost expenditure proof file into an image, extract the feature vectors and use the clustering model to classify it, and complete the cost collection based on the classification results.

Benefits of technology

Improve the efficiency of cost collection and reduce the problem of inaccurate cost collection caused by human error.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118917802A_ABST
    Figure CN118917802A_ABST
Patent Text Reader

Abstract

The invention provides an iron and steel enterprise scientific research project cost collection method and system. The method comprises the steps that a proof file related to scientific research project cost expenditure is acquired; the scientific research project expense comprises monthly labor expense of project group personnel, direct consumption material expense, inspection expense, equipment depreciation expense, equipment repair expense, power expense and fuel expense; converting the format of the proof file into an image type to form a cost expenditure image; extracting a feature vector of the expense image; based on the feature vectors, using a clustering model to complete classification of the expense images; and completing the collection of the scientific research project cost based on the classified cost and expenditure images. By using image processing and machine learning technologies, the cost collection efficiency can be improved, and the problem of inaccurate scientific research project cost collection caused by human errors can be reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of expense collection, and more specifically, relates to a method and system for collecting expenses of scientific research projects of a steel enterprise. Background Art

[0002] Research expenses are the core indicators for the recognition of high-tech enterprises. If an enterprise cannot ensure that its annual R&D investment accounts for ≥3% of its turnover, its high-tech enterprise qualification will be revoked, and the tax exemptions for additional deductions and other expenses will also be affected, which will have a serious impact on the innovation and development of the enterprise and its normal operation. Therefore, the construction of the R&D expense collection module is particularly important for enterprises to enjoy corresponding government tax exemptions, enhance corporate reputation, enhance corporate R&D driving force, and encourage innovative personnel to innovate. In the multi-process and long-process production process of steel enterprises, R&D expenses mainly include funds used to pay researchers (external and internal), production equipment and testing equipment used in the R&D process, expenses and human resources required for obtaining innovative information, and inputs into the R&D production process (material costs, fuel costs, power costs, inspection fees, etc.). For such a large enterprise, establishing a complete management system and corresponding operating rules to ensure that the cost collection is complete, accurate, and efficient is an inevitable choice to avoid risks and provide guidance for improvement. Summary of the invention

[0003] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a method and system for collecting expenses of scientific research projects of steel enterprises, so as to solve the problem of low efficiency of manual collection.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] A method for collecting expenses of scientific research projects of steel enterprises, comprising:

[0006] Step 1: Obtain supporting documents related to the expenditure of scientific research project; the expenditure of scientific research project includes: monthly labor costs of project team members, direct consumption of materials, inspection fees, equipment depreciation, equipment repair fees, power fees and fuel fees;

[0007] Step 2: Convert the format of the certification document into an image type to form an expense expenditure image;

[0008] Step 3: extracting the feature vector of the expense expenditure image;

[0009] Step 4: Using a clustering model based on the feature vector to classify the expense expenditure image;

[0010] Step 5: Complete the collection of scientific research project expenses based on the classified expense expenditure image.

[0011] Preferably, the step 3: extracting the feature vector of the expense expenditure image includes:

[0012] Using the formula:

[0013]

[0014] Extract the multi-dimensional features of the expense expenditure image; wherein G xx represents the first feature, G xy represents the second feature, G yy represents the third feature, I(x,y) represents the expense image, x represents the horizontal coordinate of the pixel point on the expense image, y represents the vertical coordinate of the pixel point on the expense image, and σ represents the standard deviation of the Gaussian filter;

[0015] A feature vector of the expense expenditure image is constructed based on the multi-dimensional features.

[0016] Preferably, the step of constructing a feature vector of the expense expenditure image based on the multi-dimensional features includes:

[0017] Using the formula:

[0018]

[0019] Construct the feature vector of the expense expenditure image; where I represents the identity matrix and * represents the convolution operation.

[0020] Preferably, the step 4: using a clustering model based on the feature vector to classify the expense expenditure image includes:

[0021] The feature vector of each expense image is input as a training sample into the deep learning model for training;

[0022] Using the loss function to continuously optimize the deep learning model to obtain a clustering model;

[0023] The target expense expenditure image is input into the clustering model to obtain the expense expenditure category to which the target expense expenditure image belongs.

[0024] The present invention also provides a system for collecting expenses of scientific research projects of steel enterprises, comprising:

[0025] The certification document acquisition module is used to obtain certification documents related to the expenditure of scientific research project expenses; the scientific research project expenses include: monthly labor costs of project team members, direct consumption material costs, inspection fees, equipment depreciation costs, equipment repair costs, power costs and fuel costs;

[0026] A format conversion module, used for converting the format of the certification document into an image type to form an expense expenditure image;

[0027] A feature vector extraction module, used to extract the feature vector of the expense expenditure image;

[0028] An image classification module, used for classifying the expense expenditure image using a clustering model based on the feature vector;

[0029] The expense collection module is used to complete the collection of scientific research project expenses based on the classified expense expenditure images.

[0030] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps in the above-mentioned method for collecting expenses of scientific research projects of a steel enterprise are implemented.

[0031] The beneficial effect of the method for collecting scientific research project expenses of a steel enterprise provided by the present invention is that compared with the prior art, the present invention can not only increase the efficiency of expense collection, but also reduce the problem of inaccurate collection of scientific research project expenses due to human errors by utilizing image processing and machine learning technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0033] Figure 1 A flow chart of a method for collecting expenses of scientific research projects of a steel enterprise provided by an embodiment of the present invention;

[0034] Figure 2 A schematic diagram of the cost collection principle for scientific research projects of a steel enterprise provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0035] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0036] See also Figure 1-2 To achieve the above purpose, the technical solution adopted by the present invention is: a method for collecting expenses of scientific research projects of steel enterprises, comprising:

[0037] Step 1: Obtain supporting documents related to the expenditure of scientific research project; the expenditure of scientific research project includes: monthly labor costs of project team members, direct consumption of materials, inspection fees, equipment depreciation, equipment repair fees, power fees and fuel fees;

[0038] In practical applications, in view of the long process, multi-process collaboration, complex equipment and a wide variety of raw materials of steel plate production lines, from the project approval, the R&D expenses generated during the operation of the project are broken down and precisely controlled, specifically:

[0039] 1. Collection of project production status

[0040] 1.1 Clarify the operation cycle of the scientific research project and the product name, process path and specific production product specification group range developed during the project operation cycle;

[0041] 1.2 Combined with the detailed product list of each production line of rare earth steel plate production line each month, the output of finished products in each scientific research project each month is counted.

[0042] 2. Collection of project labor costs

[0043] 2.1 According to the contents of the contracts of each project, sort out the specific members of the project team during the project operation phase;

[0044] 2.2 Calculate the project labor cost based on the monthly salary payment of project team members and the man-hours in the monthly work log provided by the project team. The specific calculation is:

[0045] Monthly labor cost = personnel work (salary, employee welfare, five insurances and one housing fund, enterprise annuity) * monthly project working hours / monthly total working hours

[0046] Monthly labor cost of project team members = Σ Monthly labor cost of personnel

[0047] 3. Direct consumption of material costs

[0048] Original materials of corresponding products + alloys of corresponding products + auxiliary materials of corresponding products + recycling costs of corresponding products (blast furnace return ore, gas ash, converter dust ash, recycling of heavy scrap steel, recycling of iron sheets, recycling of trimming steel scraps, recycling of zinc slag, recycling of red iron oxide - original materials)

[0049] 4. Inspection Fee

[0050] Inspection fee of corresponding products in chemical inspection center + inspection fee of corresponding products in technical center + inspection fee of corresponding products in metrology + other outsourced inspection fee of corresponding products

[0051] 5. Equipment depreciation and repair costs

[0052] Depreciation and repair costs incurred by various process equipment for products developed according to the corresponding project

[0053] 6. Power cost:

[0054] Power costs of corresponding products + costs of corresponding product recovery items (recovered steam, converter gas recovery, sintering slow cooling and hot flue gas, sintering recovery flue, blast furnace gas recovery, recovery of TRT power generation - power aspects)

[0055] 7. Fuel cost:

[0056] Corresponding product fuel cost + corresponding product recovery cost (recovery of coke powder - fuel aspect)

[0057] By continuing to collect the expenses incurred by scientific research projects in accordance with the above core technologies, it is possible to achieve accurate and complete collection of the expenses incurred by scientific research projects during the research period.

[0058] Step 2: Convert the format of the certification document into an image type to form an expense expenditure image;

[0059] In the embodiment of the present invention, the format of the certification document can be converted into an image type by taking a photo, scanning a contract, etc.

[0060] Step 3: extracting the feature vector of the expense expenditure image;

[0061] Furthermore, the step 3 comprises:

[0062] Using the formula:

[0063]

[0064] Extract the multi-dimensional features of the expense expenditure image; wherein G xx represents the first feature, G xy represents the second feature, G yy represents the third feature, I(x,y) represents the expense image, x represents the horizontal coordinate of the pixel point on the expense image, y represents the vertical coordinate of the pixel point on the expense image, and σ represents the standard deviation of the Gaussian filter;

[0065] Gaussian filters can achieve multi-scale feature extraction by adjusting the standard deviation (σ). Different σ values ​​can capture structures and textures of different sizes, making the extracted image features richer and more comprehensive.

[0066] The feature vector of the expense expenditure image is constructed based on the multi-dimensional features; wherein the feature vector calculation formula is:

[0067]

[0068] Among them, I represents the identity matrix and * represents the convolution operation.

[0069] Step 4: Using a clustering model based on the feature vector to classify the expense expenditure image;

[0070] Further, the step 4 comprises:

[0071] The feature vector of each expense image is input as a training sample into the deep learning model for training;

[0072] Using the loss function to continuously optimize the deep learning model to obtain a clustering model;

[0073] The target expense expenditure image is input into the clustering model to obtain the expense expenditure category to which the target expense expenditure image belongs.

[0074] Step 5: Complete the collection of scientific research project expenses based on the classified expense expenditure image. The collected table is as follows:

[0075] September expense collection table of an industrial application project

[0076]

[0077]

[0078] The present invention utilizes image processing and machine learning technology, which can not only increase the efficiency of cost collection, but also reduce the problem of inaccurate cost collection of scientific research projects caused by human errors.

[0079] The present invention also provides a system for collecting expenses of scientific research projects of steel enterprises, comprising:

[0080] The certification document acquisition module is used to obtain certification documents related to the expenditure of scientific research project expenses; the scientific research project expenses include: monthly labor costs of project team members, direct consumption material costs, inspection fees, equipment depreciation costs, equipment repair costs, power costs and fuel costs;

[0081] A format conversion module, used for converting the format of the certification document into an image type to form an expense expenditure image;

[0082] A feature vector extraction module, used to extract the feature vector of the expense expenditure image;

[0083] An image classification module, used for classifying the expense expenditure image using a clustering model based on the feature vector;

[0084] The expense collection module is used to complete the collection of scientific research project expenses based on the classified expense expenditure images.

[0085] Compared with the prior art, the beneficial effects of the system for collecting scientific research project expenses of a steel enterprise provided by the present invention are the same as the beneficial effects of the method for collecting scientific research project expenses of a steel enterprise described in the above technical scheme, and will not be elaborated here.

[0086] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps in the above-mentioned method for collecting expenses of scientific research projects of a steel enterprise are implemented. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present invention are the same as the beneficial effects of the method for collecting expenses of scientific research projects of a steel enterprise described in the above-mentioned technical scheme, which will not be repeated here.

[0087] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for collecting expenses of scientific research projects in steel enterprises. It is characterized in that include: Step 1: Obtain supporting documents related to the expenditure of scientific research project; The scientific research project expenses include: monthly labor costs of project team members, direct consumption of materials, inspection fees, equipment depreciation, equipment repair costs, power costs and fuel costs; Step 2: Convert the format of the certification document into an image type to form an expense expenditure image; Step 3: extracting the feature vector of the expense expenditure image; Step 4: Using a clustering model based on the feature vector to classify the expense expenditure image; Step 5: Complete the collection of scientific research project expenses based on the classified expense expenditure image.

2. A method for collecting expenses of scientific research projects of a steel enterprise as claimed in claim 1, It is characterized in that The step 3: extracting the feature vector of the expense expenditure image includes: Using the formula: Extract the multi-dimensional features of the expense expenditure image; wherein G xx represents the first feature, G xy represents the second feature, G yy represents the third feature, I(x,y) represents the expense image, x represents the horizontal coordinate of the pixel point on the expense image, y represents the vertical coordinate of the pixel point on the expense image, and σ represents the standard deviation of the Gaussian filter; A feature vector of the expense expenditure image is constructed based on the multi-dimensional features.

3. A method for collecting expenses of scientific research projects of a steel enterprise as claimed in claim 2, It is characterized in that The step of constructing a feature vector of the expense expenditure image based on the multi-dimensional features includes: Using the formula: Construct the feature vector of the expense expenditure image; where I represents the identity matrix and * represents the convolution operation.

4. A method for collecting expenses of scientific research projects of a steel enterprise as claimed in claim 3, It is characterized in that The step 4: using a clustering model based on the feature vector to complete the classification of the expense expenditure image, includes: The feature vector of each expense image is input as a training sample into the deep learning model for training; Using the loss function to continuously optimize the deep learning model to obtain a clustering model; The target expense expenditure image is input into the clustering model to obtain the expense expenditure category to which the target expense expenditure image belongs.

5. A system for collecting expenses of scientific research projects in steel enterprises. It is characterized in that include: The certification document acquisition module is used to obtain certification documents related to the expenditure of scientific research projects; The scientific research project expenses include: monthly labor costs of project team members, direct consumption of materials, inspection fees, equipment depreciation, equipment repair costs, power costs and fuel costs; A format conversion module, used for converting the format of the certification document into an image type to form an expense expenditure image; A feature vector extraction module, used to extract the feature vector of the expense expenditure image; An image classification module, used for classifying the expense expenditure image using a clustering model based on the feature vector; The expense collection module is used to complete the collection of scientific research project expenses based on the classified expense expenditure images.

6. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by the processor, the steps in the method for collecting expenses of scientific research projects of a steel enterprise as described in any one of claims 1-4 are implemented.