Construction method of pollution-reducing and carbon-reducing synergistic interaction evaluation system in dairy industry

Abstract

The invention relates to the field of environmental science and engineering, and provides a dairy industry pollution and carbon reduction synergistic interaction evaluation system construction method, which comprises the following steps: forming a comprehensive multi-dimensional data set by acquiring actual and ideal carbon footprint data of different preset dimensions, and analyzing time sequence data by using a long-short term memory model to obtain an evaluation result; and dynamically capturing a carbon emission change trend, and generating an evaluation system. The method has the beneficial effects that the generated evaluation system has flexible iteration and optimization capabilities, can be adjusted according to the actual condition of a project, ensures adaptability and timeliness, and promotes green transformation and sustainable development of the dairy industry.

Description

TECHNICAL FIELD

[0001] The application relates to the field of environmental science and engineering, and in particular to a method for constructing a dairy industry pollution reduction and carbon reduction synergistic effect evaluation system. BACKGROUND

[0002] As one of the important food industries in the world, the dairy industry inevitably produces a large amount of greenhouse gas emissions, energy consumption and waste emissions in the production process, leading to increasingly serious environmental pollution and carbon emission problems. With the increasing global pressure to reduce pollution and carbon emissions, the sustainable development of the dairy industry has become an important issue for governments and industry regulators. Therefore, establishing a scientific and effective pollution reduction and carbon reduction synergistic effect evaluation system can help the dairy industry achieve better management and optimization in the process of pollution reduction and carbon reduction, and promote the green transformation of the industry.

[0003] Existing pollution reduction and carbon reduction evaluation methods mainly focus on single indicators or partial links, and lack comprehensive evaluation in multiple dimensions. SUMMARY

[0004] The main purpose of the present application is to provide a method for constructing a dairy industry pollution reduction and carbon reduction synergistic effect evaluation system, which aims to solve the problem of lack of comprehensive evaluation in multiple dimensions.

[0005] The present application provides a method for constructing a dairy industry pollution reduction and carbon reduction synergistic effect evaluation system, comprising:

[0006] S1, obtaining actual carbon footprint data of each preset dimension of a specified pollution reduction and carbon reduction project at multiple preset time points to form an actual carbon footprint set corresponding to each preset dimension, and ideal carbon footprint data of each preset dimension at each preset time point to form an ideal carbon footprint set corresponding to each preset dimension;

[0007] S2, fusing the actual dimension carbon footprint set and the ideal dimension carbon footprint set of each preset dimension to obtain a plurality of preset dimension carbon footprint sets;

[0008] S3, inputting each element in each preset dimension carbon footprint set into a plurality of recurrent units of a preset long short-term memory model in turn to obtain a stage state output by each recurrent unit, thereby obtaining a stage state set corresponding to each preset dimension carbon footprint set;

[0009] S4, fusing each stage state set according to the preset time points to obtain a comprehensive stage state set, and labeling the comprehensive stage state set with a pollution reduction and carbon reduction score of the specified pollution reduction and carbon reduction project as a training sample;

[0010] S5, re-determine the designated pollution reduction and carbon reduction project, repeat the steps of steps S1-S4, so as to obtain a plurality of training samples;

[0011] S6, constructing a dairy industry pollution reduction and carbon reduction synergy evaluation system according to the plurality of training samples.

[0012] Further, the step S3 of sequentially inputting each element in each of the preset dimension carbon footprint set into a plurality of recurrent units of a preset long short-term memory model to obtain a stage state output by each recurrent unit, so as to obtain a stage state set corresponding to each of the preset dimension carbon footprint set, comprises:

[0013] S301, sequentially input each element in each of the preset dimension carbon footprint set into a forgetting gate, an input gate and an output gate in each recurrent unit in the long short-term memory network according to time sequence;

[0014] S302, according to the formula F t =sigmoid(W xf [X t ,Y t ]+W hf H t-1 +b f )

[0015] I t =sigmoid(W xi [X t ,Y t ]+W hi H t-1 +b i )

[0016] O t =sigmoid(W xo [X t ,Y t ]+W ho H t-1 +b o )

[0017] N t =tanh(W xn [X t ,Y t ]+W hn H t-1 +b n )

[0018] C t =F t ⊙C t-1 +I t ⊙N t

[0019] H t =O t *tanh(C t )

[0020] sequentially calculate the phase state of each cycle unit, thereby obtaining a phase state set corresponding to each of the preset dimension carbon footprint sets; wherein, F t represents the tth forgetting gate, I t represents the tth input gate, N t represents the candidate value calculated by the tth input gate, O t represents the tth output gate, C t represents the tth cell state, X t represents the actual carbon footprint data corresponding to the tth preset time point, Y t represents the ideal carbon footprint data corresponding to the tth preset time point, H t represents the phase state calculated by the tth cycle unit, W xf , W hf , W xi , W hi , W xo , W ho , W xn , W hn are preset weight parameters, respectively, b o , b i , b f , b n are preset bias parameters, respectively.

[0021] Further, the step S6 of constructing the dairy industry pollution reduction and carbon reduction synergistic evaluation system according to the plurality of training samples comprises:

[0022] S601, dividing a plurality of training samples into a training data set and a validation data set according to a preset ratio;

[0023] S602, inputting training data in the training data set into a preset classifier, and training in a supervised manner to obtain a temporary classifier;

[0024] S603, verifying the temporary classifier by verification data in the validation data set;

[0025] S604, if the verification is passed, taking the temporary classifier as a dairy industry pollution reduction and carbon reduction synergistic evaluation classifier.

[0026] Further, after the step S6 of constructing the dairy industry pollution reduction and carbon reduction synergistic evaluation system according to the plurality of training samples, the step S6 further comprises:

[0027] S701, obtain actual target carbon footprint data of each preset dimension at multiple target time points of a target pollution reduction and carbon reduction project from an initial pollution reduction and carbon reduction time to a current time, to form an actual target carbon footprint set corresponding to each preset dimension respectively, and ideal target carbon footprint data of each preset dimension at each target time point, to form an ideal target carbon footprint set corresponding to each preset dimension respectively;

[0028] S702, fuse the actual target carbon footprint set and the ideal target carbon footprint set of each preset dimension to obtain a plurality of target dimension carbon footprint sets;

[0029] S703, input each element in each target dimension carbon footprint set into a plurality of cycle units of a preset long short-term memory model in turn to obtain a target stage state output by each cycle unit, so as to obtain a target stage state set corresponding to each target dimension carbon footprint set;

[0030] S704, fuse each target stage state set according to the preset time point to obtain a comprehensive target stage state set;

[0031] S705, perform sigmoid nonlinear mapping on the comprehensive target stage state set to obtain a pollution reduction and carbon reduction index value;

[0032] S706, judge whether the pollution reduction and carbon reduction index value is less than a set threshold value;

[0033] S707, if less than the set threshold value, add a plurality of cycle units to the preset long short-term memory model;

[0034] S708, input virtual actual carbon footprint data of each preset dimension corresponding to a plurality of future target time points and corresponding ideal carbon footprint data into the added cycle units, and calculate a pollution reduction and carbon reduction index value, and continuously adjust the value of the virtual actual carbon footprint data until the calculated pollution reduction and carbon reduction index value is greater than the set threshold value, so as to obtain target actual carbon footprint data;

[0035] S709, provide advice for the target pollution reduction and carbon reduction project according to the target actual carbon footprint data and the actual target carbon footprint set.

[0036] Further, the step S709 of providing advice for the target pollution reduction and carbon reduction project according to the target actual carbon footprint data and the actual target carbon footprint set comprises:

[0037] S7091, obtain a first dimension value of each preset dimension in the actual target carbon footprint set closest to the current time, and extract a second dimension value of each preset dimension corresponding to any future time point of the target actual carbon footprint data.

[0038] S7092, calculate the difference value of the second dimension value and the first dimension value corresponding to each preset dimension;

[0039] S7093, obtain a corresponding pollution reduction and carbon reduction level according to the difference value of each preset dimension;

[0040] S7094, obtain one or more related pollution reduction and carbon reduction schemes according to the pollution reduction and carbon reduction level.

[0041] Further, the step S1 of obtaining actual carbon footprint data of each preset dimension of the specified pollution reduction and carbon reduction project at a plurality of preset time points to form an actual carbon footprint set corresponding to each preset dimension respectively comprises:

[0042] S101, obtain position information where the actual carbon footprint data is located;

[0043] S102, according to the position information, obtain actual carbon footprint data of each preset dimension of the specified pollution reduction and carbon reduction project at a plurality of preset time points through a preset crawler to form an actual carbon footprint set corresponding to each preset dimension respectively.

[0044] Further, the actual carbon footprint data of a plurality of preset dimensions includes carbon emission data in the raw material production stage, carbon emission data generated in the milk collection and transportation, carbon emission data in the dairy product processing process, carbon emission data in the production, transportation and waste process of packaging materials, and any combination of multiple data.

[0045] The present application has the following advantages: by obtaining actual and ideal carbon footprint data of different preset dimensions, a comprehensive multi-dimensional data set is formed, and a long short-term memory model is used to analyze time series data to dynamically capture the carbon emission change trend, so that the generated evaluation system has flexible iteration and optimization capability, can be adjusted according to the actual situation of the project, and ensures adaptability and timeliness, promotes the green transformation and sustainable development of the dairy product industry. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 is a flowchart of a dairy product industry pollution reduction and carbon reduction synergistic effect evaluation system construction method according to an embodiment of the present application;

[0047] Figure 2 is a structural block diagram of a computer device in an embodiment of the present application.

[0048] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0049] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below, obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the protection scope of the present application.

[0050] It should be noted that all directional indications, such as upper, lower, left, right, front, back, etc., in the embodiments of the present application are only used to explain the relative position relationship, movement condition, etc. between components in a certain posture (as shown in the drawings), and if the certain posture changes, the directional indications will also change accordingly. The connection can be direct connection or indirect connection.

[0051] The term "and / or" in this document is only used to describe the association relationship of the associated objects, which means that there can be three relationships, for example, A and B can represent three cases of existence of A alone, existence of A and B at the same time, and existence of B alone.

[0052] In addition, the description such as "first", "second" and the like in the present application is only for the purpose of description, and cannot be understood as indicating or implying the relative importance of the indicated technical features or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first", "second" can explicitly or implicitly include at least one of the features. In addition, the technical solutions of each embodiment can be combined with each other, but it must be based on the realization of ordinary skilled in the art, when the combination of technical solutions appears contradictory or unachievable, it should be considered that the combination of technical solutions does not exist, nor within the protection scope required by the present application.

[0053] Referring to Figure 1 The present application proposes a dairy industry pollution reduction and carbon reduction synergistic effect evaluation system construction method, comprising:

[0054] S1, obtaining actual carbon footprint data of each preset dimension of a specified pollution reduction and carbon reduction project at a plurality of preset time points to form an actual carbon footprint set corresponding to each preset dimension respectively, and ideal carbon footprint data of each preset dimension at each preset time point to form an ideal carbon footprint set corresponding to each preset dimension respectively;

[0055] S2, fusing the actual dimension carbon footprint set and the ideal dimension carbon footprint set of each preset dimension to obtain a plurality of preset dimension carbon footprint sets;

[0056] S3, sequentially input each element in each of the preset dimension carbon footprint set into a plurality of loop units of a preset long short-term memory model to obtain a stage state output by each loop unit, thereby obtaining a stage state set corresponding to each of the preset dimension carbon footprint set;

[0057] S4, fuse each of the stage state sets according to the preset time points to obtain a comprehensive stage state set; and label a pollution reduction and carbon reduction score of a specified pollution reduction and carbon reduction project corresponding to the comprehensive stage state set as a training sample;

[0058] S5, re-determine the specified pollution reduction and carbon reduction project, and repeat the steps of steps S1-S4, thereby obtaining a plurality of training samples;

[0059] S6, constructing a dairy industry pollution reduction and carbon reduction synergistic evaluation system according to the plurality of training samples.

[0060] As described in step S1 above, the actual carbon footprint data of the specified pollution reduction and carbon reduction project at a plurality of preset time points in each preset dimension is obtained to form an actual carbon footprint set corresponding to each preset dimension, and the ideal carbon footprint data of each preset dimension at each preset time point is obtained to form an ideal carbon footprint set corresponding to each preset dimension. These data reflect the actual emission of each dimension in the execution process of the pollution reduction and carbon reduction project in the dairy industry. The "each preset dimension" mentioned may include but is not limited to production, transportation, processing, packaging and other links, the purpose is to monitor the contribution of each link to carbon emission comprehensively. At the same time, the ideal carbon footprint data corresponding to each dimension at these time points also needs to be collected. These ideal data come from industry standards, best practices or simulation, and represent the lowest emission target of each dimension in an ideal situation.

[0061] As described in step S2 above, the actual dimension carbon footprint set and the ideal dimension carbon footprint set of each preset dimension are fused to obtain a plurality of preset dimension carbon footprint sets. The actual carbon footprint set and the ideal carbon footprint set of each preset dimension are fused. The actual data and the ideal data are combined to reveal the effectiveness and relative performance since the implementation of the pollution reduction and carbon reduction project. The fusion method can include calculating the difference between the actual carbon emission and the ideal carbon emission, and unifying the dimension by certain data processing techniques (such as normalization, standardization, etc.) in order to carry out in-depth analysis. In a specific embodiment, the fusion is only to place the data corresponding to the time points in one piece to facilitate subsequent input into the long short-term memory model.

[0062] As described in step S3, each element in each of the preset dimension carbon footprint sets is sequentially input into the multiple loop units of the preset long short-term memory model to obtain the stage state output by each loop unit, thereby obtaining a stage state set corresponding to each of the preset dimension carbon footprint sets. Each element in each of the preset dimension carbon footprint sets is sequentially input into the multiple loop units of the preset long short-term memory (LSTM) model. The time sequence characteristics and memory capacity of the LSTM model are used to extract the potential patterns in each time sequence. For each dimension carbon footprint set, the LSTM model can automatically select to retain or forget information through its internal forget gate, input gate, and output gate mechanism, thereby generating stage state output. These stage states can reflect the relationship between the current time point and the historical time points, reveal the dynamic change trend in the time series data, capture long and short-term dependencies, and output the state information of each preset time point, thereby laying a foundation for subsequent data fusion and comprehensive analysis. The preset long short-term memory model is trained by a large number of dimension carbon footprint sets and corresponding stage states. Specifically, before training begins, the parameters (weights and biases) of the LSTM model are usually initialized. In each training iteration, the input data (elements of the time series) is gradually passed through the LSTM network through the forward propagation algorithm. After forward propagation, the model produces output results (hidden states), which are compared with the true target output. An error loss function (such as mean square error or cross-entropy) is used to measure the difference between the model output and the actual result. Using the backpropagation algorithm, the weights and biases in the LSTM network are updated through the chain rule.

[0063] As described in step S4, each of the stage state sets is fused according to the preset time points to obtain a comprehensive stage state set, and the comprehensive stage state set is labeled with a pollution reduction and carbon reduction score corresponding to a specified pollution reduction and carbon reduction project as a training sample. The state information from different dimensions is combined to form a global and comprehensive state evaluation. This fusion method can be based on weighted average, maximum selection, or other aggregation algorithms to ensure that information from different dimensions can cooperatively reflect the target state. The comprehensive stage state set also needs to be labeled, and the content of the label is the pollution reduction and carbon reduction score corresponding to the specified pollution reduction and carbon reduction project as a training sample.

[0064] As described in step S5 above, the designated pollution reduction and carbon reduction project is re-determined, and steps S1-S4 are repeated to obtain multiple training samples. In order to ensure that the evaluation system constructed has wide applicability and reliability. Therefore, repeated experiments are carried out in different project backgrounds, which can help to evaluate the effect of pollution reduction and carbon reduction in different scenarios. In this process, the new project can be a completely independent pollution reduction and carbon reduction project, or an extension, refinement or adjustment of an existing project. Through continuous iteration, trial and verification, more comprehensive and accurate data samples can be accumulated to provide a richer training set for subsequent evaluation system of dairy industry pollution reduction and carbon reduction synergy.

[0065] As described in step S6 above, the dairy industry pollution reduction and carbon reduction synergy evaluation system is constructed according to the plurality of training samples. According to the plurality of training samples obtained, the dairy industry pollution reduction and carbon reduction synergy evaluation system is constructed. Through the training samples generated in the foregoing, rich dimensional information, state feedback and evaluation scores are included, which constitute the basic data source supporting the evaluation system. Then a suitable evaluation model, algorithm and index system can be selected to effectively analyze and measure the synergy effect and efficiency effect of different pollution reduction and carbon reduction projects. In the construction process, data analysis needs to be performed on the training samples to extract key influencing factors, and on this basis, appropriate evaluation standards are set. In some embodiments, the dimensional data in each training sample can also be assigned a weight, and then each preset dimension is weighted and summed to obtain the corresponding dairy industry pollution reduction and carbon reduction synergy evaluation system. Through continuous testing and optimization of the evaluation system, its accuracy and applicability in actual application can be ensured, so that the dairy industry can obtain scientific decision-making basis in the process of pollution reduction and carbon reduction, and promote the industry to move towards sustainable development.

[0066] In one embodiment, the step S3 of sequentially inputting each element in each of the preset dimension carbon footprint sets into a plurality of recurrent units of a preset long short-term memory model to obtain a stage state output by each recurrent unit, thereby obtaining a stage state set corresponding to each of the preset dimension carbon footprint sets, comprises:

[0067] S301, sequentially input each element in each of the preset dimension carbon footprint sets into a forget gate, an input gate and an output gate in each recurrent unit in the long short-term memory network according to time sequence;

[0068] S302, according to the formula F t = sigmoid(W xf [X t , Y t ]+W hf H t-1 +b f )

[0069] I t = sigmoid(W xi [X t , Y t ] + W hi H t-1 + b i )

[0070] O t = sigmoid(W xo [X t , Y t ] + W ho H t-1 + b o )

[0071] N t = tanh(W xn [X t , Y t ] + W hn H t-1 + b n )

[0072] C t = F t ⊙C t-1 + I t ⊙N t

[0073] H t = O t *tanh(C t )

[0074] The phase state of each cycle unit is calculated in turn, so as to obtain a phase state set corresponding to each preset dimension carbon footprint set; wherein, F t represents the tth forgetting gate, I t represents the tth input gate, N t represents the candidate value calculated by the tth input gate, O t represents the tth output gate, C t represents the tth cell state, X t represents the actual carbon footprint data corresponding to the tth preset time point, Y t represents the ideal carbon footprint data corresponding to the tth preset time point, H t represents the phase state calculated by the tth cycle unit, W xf , W hf , W xi , W hi , W xo , W ho , W xn , W hn are preset weight parameters, respectively, and bo , b i , b f , b n are preset bias parameters, respectively.

[0075] As described in steps S301-S302, each element in each of the preset dimension carbon footprint sets is sequentially input into the forget gate, the input gate, and the output gate of each recurrent unit in the long short-term memory network in time sequence, thereby realizing the calculation of the stage state of each recurrent unit. It should be noted that the number of recurrent units is the same as the number of preset time points.

[0076] In one embodiment, the step S6 of constructing a dairy industry pollution reduction and carbon reduction synergy evaluation system according to the plurality of training samples comprises:

[0077] S601, dividing a plurality of training samples into a training data set and a validation data set according to a preset ratio:

[0078] S602, inputting training data in the training data set into a preset classifier and training in a supervised manner to obtain a temporary classifier;

[0079] S603, verifying the temporary classifier by verification data in the verification data set;

[0080] S604, if the verification is passed, taking the temporary classifier as a dairy industry pollution reduction and carbon reduction synergy evaluation classifier.

[0081] As described in steps S601-S604, first, a plurality of training samples need to be divided into a training data set and a verification data set according to a preset ratio. A plurality of training samples generated before are collected, which contain comprehensive stage states and corresponding pollution reduction and carbon reduction scores for different pollution reduction and carbon reduction projects. According to the project requirements and sample size, the division ratio of the training data set and the verification data set is set. For example, it is set to 70:30 or 80:20. In order to avoid sample bias, samples should be randomly extracted and divided into the training data set and the verification data set. Ensure that each data set can represent the characteristics of the overall sample, thereby improving the effectiveness of subsequent model training and verification.

[0082] The data in the training dataset is input into the preset classifier, and supervised training is performed to obtain a temporary classifier. A suitable machine learning classifier is selected according to the task requirements, such as decision tree, random forest, support vector machine (SVM), or neural network, etc. These models will be used to predict the pollution reduction and carbon reduction scores. The input features (such as the set of stage states) and labels (pollution reduction and carbon reduction scores) in the prepared training dataset are input into the classifier for training. Using supervised learning, the model gradually learns how to effectively predict the output through the relationship between the input features and the corresponding labels. The algorithm adjusts the parameters of the model by optimizing the loss function, so that the model's prediction results are as close to the actual labels as possible. During the training process, cross-validation techniques can be used to further optimize the model parameters to ensure the stability and generalization ability of the model on different data subsets. The training results fed back by the model improve the accuracy of the model. After training is completed, a temporary classifier is obtained, and the temporary classifier is verified by the data in the validation dataset. The goal of this step is to evaluate the accuracy and generalization ability of the model. The input features (set of stage states) in the validation dataset are input into the trained temporary classifier for prediction. The predicted results of the classifier are compared with the actual labels (pollution reduction and carbon reduction scores) in the validation dataset, and the performance of the model is evaluated by calculating the precision, recall, F1 score, etc. If the validation result is not satisfactory, further optimization strategies may be needed, such as adjusting feature selection, retraining, or using other algorithms for comparative analysis. In addition, data augmentation and other methods can be considered to improve the generalization performance of the model. If the verification is passed, the temporary classifier is used as the dairy industry pollution reduction and carbon reduction synergy evaluation classifier.

[0083] In one embodiment, after the step S6 of constructing the dairy industry pollution reduction and carbon reduction synergy evaluation system according to the plurality of training samples, the method further comprises:

[0084] S701, obtaining actual target carbon footprint data of each preset dimension at a plurality of target time points within an initial pollution reduction and carbon reduction time to a current time of a target pollution reduction and carbon reduction project, to form an actual target carbon footprint set corresponding to each preset dimension respectively, and ideal target carbon footprint data of each preset dimension at each target time point, to form an ideal target carbon footprint set corresponding to each preset dimension respectively;

[0085] S702, fusing the actual target carbon footprint set and the ideal target carbon footprint set of each preset dimension to obtain a plurality of target dimension carbon footprint sets;

[0086] S703, input each element in each of the target dimension carbon footprint set into a plurality of recurrent units of a preset long short-term memory model in turn to obtain a target stage state output by each recurrent unit, so as to obtain a target stage state set corresponding to each of the target dimension carbon footprint set;

[0087] S704, fuse each of the target stage state set according to the preset time point to obtain a comprehensive target stage state set;

[0088] S705, perform sigmoid nonlinear mapping on the comprehensive target stage state set to obtain a pollution reduction and carbon reduction index value;

[0089] S706, judge whether the pollution reduction and carbon reduction index value is less than a set threshold value;

[0090] S707, if less than the set threshold value, add a plurality of recurrent units to the preset long short-term memory model;

[0091] S708, input virtual actual carbon footprint data of each preset dimension corresponding to a plurality of future target time points and corresponding ideal carbon footprint data to the added recurrent unit, calculate a pollution reduction and carbon reduction index value, and continuously adjust the value of the virtual actual carbon footprint data until the calculated pollution reduction and carbon reduction index value is greater than the set threshold value, so as to obtain target actual carbon footprint data;

[0092] S709, provide a suggestion for the target pollution reduction and carbon reduction project according to the target actual carbon footprint data and the actual target carbon footprint set.

[0093] As described above, the steps S701-S704 achieve the acquisition of the comprehensive target stage state set of the target pollution reduction and carbon reduction project. The specific acquisition method can refer to the acquisition method of the comprehensive stage state set. The method of acquiring the comprehensive target stage state set is the same as that of acquiring the comprehensive stage state set, which will not be described here.

[0094] As described in steps S705-S709 above, the comprehensive target phase state set is subjected to Sigmoid nonlinear mapping to generate a pollution reduction and carbon reduction index value. The Sigmoid function is a commonly used nonlinear activation function. By applying this function, the values in the comprehensive target phase state set can be compressed into the interval (0, 1), thereby generating a standardized index value. Through nonlinear mapping, the generation of the pollution reduction and carbon reduction index value provides a digital form of expression for subsequent classification and decision-making. This index value can better reflect the overall evaluation output of the current model for pollution reduction and carbon reduction projects. It is determined whether the obtained pollution reduction and carbon reduction index value is less than a set threshold value. According to business requirements and industry standards, a threshold value for the pollution reduction and carbon reduction index is set. Generally, this threshold value represents the basic compliance standard of the project and is the benchmark for evaluating effectiveness. The calculated pollution reduction and carbon reduction index value is compared with the set threshold value. According to the characteristics of the industry and the requirements of the project, it is determined whether the current index performance meets the expectations or needs to be adjusted. If the pollution reduction and carbon reduction index value is greater than the set threshold value, it generally indicates that the implementation effect of the project in pollution reduction and carbon reduction is good, and the existing strategy and execution mode can be continued.

[0095] If the pollution reduction and carbon reduction index value is less than the set threshold value, a plurality of recurrent units are added to the preset long short-term memory model to enhance the expression ability of the model. The virtual actual carbon footprint data of each preset dimension and the corresponding ideal carbon footprint data of the future target time point are input into the added recurrent units, and the pollution reduction and carbon reduction index value is calculated. This step is crucial for predicting future performance. For the future target time point, virtual actual carbon footprint data and corresponding ideal carbon footprint data are generated. The generated virtual carbon footprint data is input into the expanded LSTM network for new prediction and calculation. If the currently calculated pollution reduction and carbon reduction index value is still less than the set threshold value, the value of the virtual actual carbon footprint data is continuously adjusted until the calculated pollution reduction and carbon reduction index value is greater than the set threshold value.

[0096] According to the obtained target actual carbon footprint data, in combination with the actual target carbon footprint set, corresponding suggestions are generated. The suggestion content can include further optimization measures, improvement of project implementation plan, or re-evaluation of existing strategies. The gap between the actual carbon footprint data and the ideal carbon footprint data is analyzed to identify potential problem areas and propose corresponding optimization measures. This can cover aspects such as production process, raw material selection, energy use efficiency, etc. The proposed suggestions should be combined with relevant industry standards and best practices to ensure the reasonableness and implementability of the suggestions. Successful cases of other enterprises can be explored for reference.

[0097] In one embodiment, the step S709 of providing suggestions for the target pollution reduction and carbon reduction project according to the target actual carbon footprint data and the actual target carbon footprint set comprises:

[0098] S7091, obtaining a first dimension value of each preset dimension closest to the current time in the actual target carbon footprint set, and extracting a second dimension value of each preset dimension corresponding to the target actual carbon footprint number at any future time point;

[0099] S7092, calculating the difference between the second dimension value and the first dimension value corresponding to each preset dimension;

[0100] S7093, obtaining the corresponding pollution reduction and carbon reduction level according to the difference of each preset dimension;

[0101] S7094, obtaining one or more related pollution reduction and carbon reduction schemes according to the pollution reduction and carbon reduction level.

[0102] As described in steps S7091-S7094, the first dimension value of each preset dimension closest to the current time in the actual target carbon footprint set is obtained, and the second dimension value of each preset dimension corresponding to the target actual carbon footprint number at any future time point is extracted. The difference between the second dimension value and the first dimension value corresponding to each preset dimension is calculated. Through comparative analysis, the field that needs to be improved is found in the evaluation of the pollution reduction and carbon reduction project. The difference obtained by calculation is the second dimension minus the first dimension, which is generally negative, so the difference is generally the absolute value. If it is positive, it means that the pollution reduction and carbon reduction of this dimension is very good, and it can be maintained without adjustment. According to the difference of each preset dimension, the corresponding pollution reduction and carbon reduction level is obtained. The difference is converted into an operational grading system to provide a basis for specific pollution reduction and carbon reduction schemes. According to industry standards, historical data or company internal regulations, different pollution reduction and carbon reduction levels are set. For example, it can be divided into "excellent", "good", "medium" and "poor" levels, each of which corresponds to a specific difference range.

[0103] The difference of each preset dimension is mapped to the corresponding pollution reduction and carbon reduction level through the designed level division. For example, if the difference is greater than a certain threshold, it can be determined as "needs improvement", and if it is within a certain normal range, it is set as "meets the standard". According to the pollution reduction and carbon reduction level, one or more related pollution reduction and carbon reduction schemes are obtained. Specifically, a corresponding pollution reduction and carbon reduction scheme library can be established in advance, and these schemes should be developed according to different industry standards, practical experience and technical development. The database should include specific measures and improvement suggestions related to each pollution reduction and carbon reduction level. In order to cope with the characteristics and background of different dimensions, one or more candidate schemes can be provided to ensure that the enterprise can select the most suitable scheme according to the actual situation. This can improve the flexibility and adaptability of pollution reduction and carbon reduction.

[0104] In one embodiment, the step S1 of obtaining actual carbon footprint data of a designated pollution reduction and carbon reduction project at a plurality of preset time points in each preset dimension to form an actual carbon footprint set corresponding to each preset dimension respectively comprises:

[0105] S101, acquire position information where the actual carbon footprint data is located;

[0106] S102, according to the position information, acquire actual carbon footprint data of each preset dimension of the specified pollution reduction and carbon reduction project at multiple preset time points through a preset crawler to form an actual carbon footprint set corresponding to each preset dimension respectively.

[0107] As described in steps S101-S102, the position information where the actual carbon footprint data is located is the IP address of the database where the actual carbon footprint data is located. The rules of the crawler are set according to multiple preset dimensions (such as production process, transportation process, energy consumption, etc.) to ensure that the actual carbon footprint data of each dimension can be completely captured.

[0108] In one embodiment, the actual carbon footprint data of the multiple preset dimensions includes any combination of carbon emission data of raw material production stage, carbon emission data generated during milk collection and transportation, carbon emission data during dairy product processing, carbon emission data during production, transportation and disposal of packaging materials. Among them, the carbon emission data of the raw material production stage generally covers the greenhouse gases emitted during the production process of various raw materials (such as milk, feed, etc.) used in dairy product production. These emissions are usually caused by the production and transportation of fertilizers and feed. The carbon emission data generated during milk collection and transportation is the carbon emission data generated during the transportation process of milk from the farm to the processing plant, including the carbon emission generated in each link from the farm to the dairy product processing plant, which is usually related to the type of transportation tool (such as truck, tank truck, etc.), transportation distance and frequency, etc. The carbon emission data during dairy product processing is the carbon emission data generated during the processing of dairy products involving different processes (such as pasteurization, product making, and sub-packaging, etc.), which consumes energy and generates corresponding carbon emissions. Mainly from the energy consumption of production machinery and equipment, chemical reactions of process operation, etc. The carbon emission data during the production, transportation and disposal of packaging materials is the carbon emission data generated during the packaging process of dairy products, including the selection of packaging materials, the production process, the transportation of packaging materials from the factory to the consumer market, and the carbon emission generated during the final disposal process (such as recycling, incineration or landfill, etc.).

[0109] The beneficial effects of the present application: by acquiring actual and ideal carbon footprint data of different preset dimensions, forming a comprehensive multi-dimensional data set, and then using a long short-term memory (LSTM) model to analyze time series data, the dynamic trend of carbon emission is captured, so that the generated evaluation system has flexible iteration and optimization capability, which can be adjusted according to the actual situation of the project, ensuring adaptability and timeliness, and promoting the green transformation and sustainable development of the dairy product industry.

[0110] Referring to Figure 2 In the embodiments of the present application, a computer device is also provided, which can be a server, and the internal structure of the computer device can be as shown in Figure 2 The computer device includes a processor, a memory, a network interface and a database connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The database of the computer device is configured to store various carbon footprint data and the like. The network interface of the computer device is configured to communicate with an external terminal through a network connection. The computer program, when executed by the processor, can implement the dairy industry pollution reduction and carbon reduction synergistic effect evaluation system construction method described in any of the above embodiments.

[0111] Those skilled in the art can understand that Figure 2 The structure shown in the above is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied.

[0112] The embodiments of the present application also provide a computer readable storage medium having a computer program stored thereon, and the computer program, when executed by the processor, can implement the dairy industry pollution reduction and carbon reduction synergistic effect evaluation system construction method described in any of the above embodiments.

[0113] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiment methods. Any reference to memory, storage, databases, or other media in this application and in examples provided herein, unless specifically stated otherwise, can include non-volatile and / or volatile memory. Non-volatile memory can include, for example, read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include, for example, random access memory (RAM), or external cache memory. As an illustration and not a limitation, RAM can be available in many forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), Rambus DRAM (RDRAM), direct Rambus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0114] It should be noted that the terms "comprising", "including", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, a device, an article or a method that comprises a list of elements does not include only those elements recited, but can also include other elements not expressly listed or inherent to such process, device, article or method. Without further limitation, an element defined by the phrase "comprising a" does not exclude the presence of additional identical elements in the process, device, article or method that includes the element.

[0115] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Those skilled in the art can make various modifications and changes to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the scope of the claims of the present application.

Claims

1. A method for constructing an evaluation system for synergistic efficiency improvement in pollution reduction and carbon reduction in the dairy industry, characterized in that, include: S1. Obtain the actual carbon footprint data of the specified pollution reduction and carbon reduction project at multiple preset time points for each preset dimension, so as to form the actual carbon footprint set corresponding to each preset dimension, and the ideal carbon footprint data of each preset dimension at each preset time point, so as to form the ideal carbon footprint set corresponding to each preset dimension. S2. Merge the actual carbon footprint set of each preset dimension with the ideal carbon footprint set to obtain multiple preset dimension carbon footprint sets. S3. Input each element in each preset dimension carbon footprint set into multiple loop units of the preset long short-term memory model in sequence to obtain the stage state output by each loop unit, thereby obtaining the stage state set corresponding to each preset dimension carbon footprint set. S4. Merge each set of stage states according to the preset time points to obtain a comprehensive set of stage states; The pollution reduction and carbon reduction scores of the corresponding designated pollution reduction and carbon reduction projects are labeled on the integrated stage state set as a training sample. S5. Re-determine the designated pollution reduction and carbon reduction projects, and repeat steps S1-S4 to obtain multiple training samples. S6. Construct an evaluation system for synergistic efficiency improvement in pollution reduction and carbon reduction in the dairy industry based on the multiple training samples.

2. The method for constructing an evaluation system for synergistic efficiency improvement in pollution reduction and carbon reduction in the dairy industry as described in claim 1, characterized in that, Step S3, which involves sequentially inputting each element in each of the preset dimension carbon footprint sets into multiple loop units of a preset long short-term memory model to obtain the stage state output by each loop unit, thereby obtaining the stage state set corresponding to each preset dimension carbon footprint set, includes: S301. Input each element in each preset dimension carbon footprint set into the forget gate, input gate, and output gate of each recurrent unit in the long short-term memory network in chronological order. S302, According to formula F t =sigmoid(W xf [X t ,Y t ]+W hf H t-1 +b f ) I t =sigmoid(W xi [X t ,Y t ]+W hi H t-1 +b i ) O t =sigmoid(W xo [X t ,Y t ]+W ho H t-1 +b o ) N t =tanh(W xn [X t ,Y t ]+W hn H t-1 +b n ) C t =F t ⊙C t-1 +I t ⊙N t H t =O t *fishy( C t) The stage states of each loop unit are calculated sequentially to obtain the stage state set corresponding to each preset dimension carbon footprint set; where F t Let I represent the t-th forget gate. t N represents the t-th input gate. t This represents the candidate value calculated by the t-th input gate, O t Let C represent the t-th output gate. t Let X represent the state of the t-th cell. t Y represents the actual carbon footprint data corresponding to the t-th preset time point. t H represents the ideal carbon footprint data corresponding to the t-th preset time point. t W represents the stage state calculated by the t-th loop unit. xf W hf W xi W hi W xo W ho W xn W hn These are the preset weight parameters, b o b i b f b n These are the preset bias parameters.

3. The method for constructing an evaluation system for synergistic efficiency improvement in pollution reduction and carbon reduction in the dairy industry as described in claim 1, characterized in that, Step S6, which involves constructing a collaborative evaluation system for pollution reduction and carbon reduction in the dairy industry based on the multiple training samples, includes: S601. Divide the multiple training samples into a training dataset and a validation dataset according to a preset ratio; S602. Input the training data in the training dataset into a preset classifier and train it in a supervised manner to obtain a temporary classifier. S603. Validate the temporary classifier using the validation data in the validation dataset; S604. If the verification is successful, the temporary classifier shall be used as the evaluation classifier for the synergistic effect of pollution reduction and carbon reduction in the dairy industry.

4. The method for constructing an evaluation system for synergistic efficiency improvement in pollution reduction and carbon reduction in the dairy industry as described in claim 1, characterized in that, Following step S6, which involves constructing a collaborative evaluation system for pollution reduction and carbon reduction in the dairy industry based on the multiple training samples, the following steps are also included: S701. Obtain actual target carbon footprint data for each preset dimension at multiple target time points from the initial time of pollution reduction and carbon reduction project to the present time, so as to form the actual target carbon footprint set corresponding to each preset dimension, and the ideal target carbon footprint data for each preset dimension at each target time point, so as to form the ideal target carbon footprint set corresponding to each preset dimension. S702. Merge the actual target carbon footprint set of each preset dimension with the ideal target carbon footprint set to obtain multiple target dimension carbon footprint sets. S703. Input each element in each target dimension carbon footprint set into multiple loop units of a preset long short-term memory model in sequence to obtain the target stage state output by each loop unit, thereby obtaining the target stage state set corresponding to each target dimension carbon footprint set. S704. Merge each target stage state set according to the preset time point to obtain a comprehensive target stage state set; S705. Perform a sigmoid nonlinear mapping on the set of states of the comprehensive target stage to obtain pollution reduction and carbon reduction index values. S706. Determine whether the pollution reduction and carbon reduction index value is less than the set threshold. S707. If the value is less than the set threshold, then multiple loop units are added to the preset long short-term memory model. S708. Input the virtual actual carbon footprint data of each preset dimension corresponding to multiple future target time points and the corresponding ideal carbon footprint data into the added loop unit, calculate the pollution reduction and carbon reduction index value, and continuously adjust the value of the virtual actual carbon footprint data until the calculated pollution reduction and carbon reduction index value is greater than the set threshold, thereby obtaining the target actual carbon footprint data. S709. Provide recommendations for the target pollution reduction and carbon reduction projects based on the actual carbon footprint data of the target and the set of actual target carbon footprints.

5. The method for constructing an evaluation system for synergistic efficiency improvement in pollution reduction and carbon reduction in the dairy industry as described in claim 4, characterized in that, Step S709, which provides recommendations for the target pollution reduction and carbon reduction project based on the target's actual carbon footprint data and the actual target carbon footprint set, includes: S7091. Obtain the first dimension value of each preset dimension that is closest to the current time in the actual target carbon footprint set, and extract the second dimension value of each preset dimension corresponding to any future time point of the actual target carbon footprint. S7092. Calculate the difference between the second dimension value and the first dimension value corresponding to each preset dimension; S7093. Obtain the corresponding pollution reduction and carbon reduction level based on the difference between each preset dimension; S7094. Obtain one or more relevant pollution reduction and carbon reduction schemes based on the pollution reduction and carbon reduction level.

6. The method for constructing an evaluation system for synergistic efficiency improvement in pollution reduction and carbon reduction in the dairy industry as described in claim 1, characterized in that, Step S1, which involves obtaining actual carbon footprint data of a specified pollution reduction and carbon reduction project at multiple preset time points for each preset dimension to form a set of actual carbon footprints corresponding to each preset dimension, includes: S101. Obtain the location information of the actual carbon footprint data; S102. Based on the location information, the actual carbon footprint data of the specified pollution reduction and carbon reduction project at multiple preset time points and in various preset dimensions are obtained through a preset crawler, so as to form a set of actual carbon footprints corresponding to each preset dimension.

7. The method for constructing an evaluation system for synergistic efficiency improvement in pollution reduction and carbon reduction in the dairy industry as described in claim 1, characterized in that, The actual carbon footprint data of the multiple preset dimensions includes various combinations of data from the carbon emission data of the raw material production stage, the carbon emission data generated by milk collection and transportation, the carbon emission data of the dairy product processing process, and the carbon emission data of the production, transportation and disposal of packaging materials.

Images