Carbon emission monitoring method and system based on block chain and neural network

Abstract

The invention discloses a carbon emission monitoring method and system based on a block chain and a neural network. The method comprises the following steps: each participating enterprise party and supervisor deploys a block chain node and a business service system; deploying a smart contract on the block chain; each participating enterprise party receives and stores activity data of the respective enterprise through the respective business service system, sends the activity data to the block chain, and calls an intelligent contract to calculate and obtain a carbon emission index calculation result; the supervision party obtains activity data and carbon emission index calculation results of all participating enterprise parties, and performs model training and testing based on a long-short-term memory neural network to obtain a carbon emission prediction model; and each participating enterprise predicts future carbon emission according to local activity data and a carbon emission index calculation result based on the latest carbon emission prediction model. According to the method, the block chain technology and the neural network model are utilized, automatic monitoring and processing of the carbon emission data are realized, and the accuracy of the data can be ensured.

Description

Technical Field

[0001] The present invention relates to the field of blockchain technology, and in particular to a carbon emission monitoring method and system based on blockchain and neural network. Background Art

[0002] Existing carbon emission monitoring methods mainly rely on manual data collection and manual calculation, which is not only time-consuming and labor-intensive, but also prone to errors. With the development of blockchain technology and neural network technology, how to use these advanced technologies to automatically and accurately monitor carbon emissions has become an urgent problem to be solved.

[0003] Currently, no effective solution has been proposed to address the problems of low processing efficiency and low accuracy in the existing technology of manually collecting data and manually calculating carbon emissions. Summary of the invention

[0004] The embodiments of the present invention provide a carbon emission monitoring method and system based on blockchain and neural network to solve the problems of low processing efficiency and low accuracy in the prior art of manually collecting data and manually calculating carbon emissions.

[0005] To achieve the above-mentioned purpose, on the one hand, the present invention provides a carbon emission monitoring method based on blockchain and neural network, which method includes: S1, each participating enterprise and regulator deploys blockchain nodes and business service systems; S2, deploys smart contracts on the blockchain; S3, each participating enterprise receives and stores the activity data of their respective enterprises through their respective business service systems; each business service system sends its own activity data to the blockchain, and calls the smart contract to perform consensus on activity data storage, calculate carbon emission indicators on the activity data to obtain carbon emission indicator calculation results, store consensus on the carbon emission indicator calculation results and save them in the local database of each enterprise, and judge according to the carbon emission indicator calculation results Whether to issue an early warning notification, if so, the early warning notification consensus is communicated to the regulator; S4, the regulator obtains the activity data and carbon emission index calculation results of all participating enterprises, and trains and tests the model based on the long short-term memory neural network to obtain the carbon emission prediction model; the configuration file of the carbon emission prediction model is saved to obtain the storage address, and the smart contract is called to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus; S5, each participating enterprise saves the storage address and version number of the latest carbon emission prediction model to the local database, so that each participating enterprise can predict the future carbon emissions of each participating enterprise based on the local activity data and carbon emission index calculation results based on the latest carbon emission prediction model.

[0006] Optionally, the smart contract includes: a data storage contract, a carbon index calculator contract, a neural network model management contract, and an early warning contract.

[0007] Optionally, S3 includes: each participating enterprise party connects the activity data of each enterprise to its own business service system through the REST API interface, and saves it in the local database of each enterprise; each business service system communicates with the blockchain through the blockchain SDK, uploads its own activity data to the blockchain, and calls the data evidence contract for evidence storage, and distributes it to the regulator of the blockchain through the consensus mechanism; each participating enterprise party calls the carbon index calculator contract to calculate the carbon emission index of its own activity data to obtain its own carbon emission index calculation result; each participating enterprise party calls the data evidence contract to store its own carbon emission index calculation result, and distributes it to the regulator of the blockchain through the consensus mechanism, and saves its own carbon emission index calculation result in the local database of each enterprise; each participating enterprise party calls the early warning contract to compare its own carbon emission index calculation result with the preset index threshold. If the carbon emission index calculation result is greater than the preset index threshold, an early warning notification is issued, and the early warning notification consensus is sent to the regulator based on the notification mechanism of the early warning contract. Otherwise, no early warning notification is issued.

[0008] Optionally, S4 includes: the regulator obtains the activity data and carbon emission index calculation results of all participating corporate parties based on the data notarization contract; classifies the activity data and carbon emission index calculation results of all participating corporate parties, and standardizes the classified data; divides the standardized classified data into a training set and a test set; iteratively trains the training set based on the long short-term memory neural network to update the long short-term memory neural network, and predicts the test set through the long short-term memory neural network after the current iteration to obtain a prediction result; if the root mean square error between the prediction result and the true result is not within a preset range, the training set is iteratively trained for the next time based on the long short-term memory neural network after the current iteration until the root mean square error between the prediction result and the true result is within the preset range, stops the iterative training, and uses the long short-term memory neural network obtained by the last iteration as the carbon emission prediction model; saves the configuration file of the carbon emission prediction model to the cloud storage service to obtain the storage address, and calls the neural network model management contract to upload the storage address and version number of the carbon emission prediction model to the chain for notarization consensus.

[0009] Optionally, when the carbon emission prediction model is updated, the configuration file of the updated carbon emission prediction model is saved to the cloud storage service to obtain the storage address, and the neural network model management contract is called to upload the storage address and version number of the updated carbon emission prediction model to the chain for evidence consensus.

[0010] Optionally, S5 includes: each participating enterprise saves the storage address and version number of the latest carbon emission prediction model to the local database; each participating enterprise classifies the activity data and carbon emission index calculation results in the local database, and standardizes the classified data; and uses the standardized classified data through the latest carbon emission prediction model to predict the future carbon emissions of each participating enterprise.

[0011] On the other hand, the present invention provides a carbon emission monitoring system based on blockchain and neural network, which includes: a node and system deployment unit, which is used for each participating enterprise and regulator to deploy blockchain nodes and business service systems; a smart contract deployment unit, which is used to deploy smart contracts on the blockchain; a storage and calculation unit, which is used for each participating enterprise to receive and store the activity data of each enterprise through their respective business service systems; each business service system sends its own activity data to the blockchain, and calls the smart contract to perform activity data storage consensus in turn, calculate the carbon emission index of the activity data to obtain the carbon emission index calculation result, store the carbon emission index calculation result consensus and save it in the local database of each enterprise, and calculate the carbon emission index calculation result according to the carbon emission index calculation result. The result determines whether to issue an early warning notification. If so, the early warning notification consensus is communicated to the regulator; the model training unit is used by the regulator to obtain the activity data and carbon emission index calculation results of all participating enterprises, and to train and test the model based on the long short-term memory neural network to obtain the carbon emission prediction model; the configuration file of the carbon emission prediction model is saved to obtain the storage address, and the smart contract is called to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus; the prediction unit is used by each participating enterprise to save the storage address and version number of the latest carbon emission prediction model to the local database, so that each participating enterprise can predict the future carbon emissions of each participating enterprise based on the local activity data and carbon emission index calculation results based on the latest carbon emission prediction model.

[0012] Optionally, the smart contract includes: a data storage contract, a carbon index calculator contract, a neural network model management contract, and an early warning contract.

[0013] Optionally, the evidence storage and calculation unit includes: a receiving storage subunit for each participating enterprise party to receive the information through REST The API interface connects the activity data of each enterprise to its own business service system and saves it in the local database of each enterprise; the activity data evidence consensus sub-unit is used for each business service system to communicate with the blockchain through the blockchain SDK, upload their own activity data to the blockchain, call the data evidence contract for evidence, and distribute it to the regulator of the blockchain through the consensus mechanism; the calculation sub-unit is used for each participating enterprise to call the carbon index calculator contract to calculate the carbon emission index of their own activity data to obtain their own carbon emission index calculation results; the carbon emission index calculation result storage sub-unit is used for each participating enterprise to call the data evidence contract to store their own carbon emission index calculation results, distribute them to the regulator of the blockchain through the consensus mechanism, and save their own carbon emission index calculation results in the local database of each enterprise; the early warning sub-unit is used for each participating enterprise to call the early warning contract to compare their own carbon emission index calculation results with the preset index threshold. If the carbon emission index calculation result is greater than the preset index threshold, an early warning notification is issued, and the early warning notification consensus is sent to the regulator based on the notification mechanism of the early warning contract. Otherwise, no early warning notification is issued.

[0014] Optionally, the model training unit includes: an acquisition subunit, which is used by the regulator to obtain the activity data and carbon emission index calculation results of all participating corporate parties based on the data notarization contract; a first standardization subunit, which is used to classify the activity data and carbon emission index calculation results of all participating corporate parties, and standardize the classified data; a model training subunit, which is used to divide the classified data after standardization into a training set and a test set; iteratively train the training set based on the long short-term memory neural network to update the long short-term memory neural network, and predict the test set through the long short-term memory neural network after the current iteration to obtain a prediction result; if the root mean square error between the predicted result and the true result is not within a preset range, perform the next iterative training on the training set based on the long short-term memory neural network after the current iteration until the root mean square error between the predicted result and the true result is within the preset range, stop the iterative training, and use the long short-term memory neural network obtained from the last iteration as the carbon emission prediction model; a model information notarization consensus subunit, which is used to save the configuration file of the carbon emission prediction model to the cloud storage service to obtain the storage address, and call the neural network model management contract to upload the storage address and version number of the carbon emission prediction model to the chain for notarization consensus.

[0015] Beneficial effects of the present invention:

[0016] The present invention provides a carbon emission monitoring method and system based on blockchain and neural network, wherein the method utilizes blockchain technology and neural network model to realize automatic monitoring and processing of carbon emission data. The entire process from data access, evidence storage, calculation to early warning notification and model training can be automatically executed, reducing the need for manual operation and greatly improving the automation level of the system; through the carbon index calculator contract, the system can automatically calculate accurate carbon emission indicators. At the same time, the long short-term memory neural network is used to train and predict historical data, which improves the prediction accuracy of future carbon emission trends. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flowchart of a carbon emission monitoring method based on blockchain and neural network provided by an embodiment of the present invention;

[0018] Figure 2 is a flow chart of data storage and calculation provided by an embodiment of the present invention;

[0019] Figure 3 is a flow chart of model training provided by an embodiment of the present invention;

[0020] Figure 4 is a flow chart of future carbon emissions prediction provided by an embodiment of the present invention;

[0021] Figure 5 It is a structural schematic diagram of a carbon emission monitoring system based on blockchain and neural network provided by an embodiment of the present invention;

[0022] Figure 6 is a schematic diagram of the structure of the evidence storage and calculation unit provided in an embodiment of the present invention;

[0023] Figure 7 is a schematic diagram of the structure of a model training unit provided in an embodiment of the present invention;

[0024] Figure 8 It is a schematic diagram of the structure of the prediction unit provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0026] Figure 1 A carbon emission monitoring method based on blockchain and neural network is provided in an embodiment of the present invention. Figure 1As shown, the method includes:

[0027] S1. Each participating enterprise and regulator deploys blockchain nodes and business service systems;

[0028] Each participating enterprise will deploy its own blockchain node. The purpose of deploying blockchain nodes is: 1. Achieving data consensus: Each node participates in the consensus mechanism of the blockchain to ensure that all nodes reach a consensus on the data on the chain; 2. Data storage and verification: Each node saves and verifies all transaction records on the chain to ensure the security and non-tamperability of the data.

[0029] Each participating enterprise will deploy its own business service system, because enterprise users need to access blockchain-related information, query data storage, query early warning notifications, query carbon index data, etc. through the business service system. Therefore, each enterprise needs an independent business service system to provide services for its internal users.

[0030] Regulators deploying blockchain nodes can participate in the data consensus process of the entire blockchain network, which means that the regulator's nodes can jointly verify and record all data on the chain with other corporate nodes to ensure the integrity and reliability of the data.

[0031] The deployment of a business service system by regulators can provide regulators with a centralized platform for querying carbon emission data, early warning information, activity data, etc. of all enterprises. In this way, regulators can easily access and manage data in the entire blockchain network; further, the business service system can also help regulators receive and process early warning notifications from the blockchain network, so as to take timely actions to avoid potential risks.

[0032] S2. Deploy smart contracts on the blockchain;

[0033] In an optional implementation, the smart contract includes: a data storage contract, a carbon index calculator contract, a neural network model management contract, and an early warning contract.

[0034] The main function of the data storage contract is to store the enterprise's data (such as activity data and carbon emission index calculation results), that is, to write these data into the blockchain to ensure the data is tamper-proof and transparent;

[0035] The carbon index calculator contract is used to perform business calculations on the activity data of the enterprise. Specifically, it automatically calculates the carbon emission index calculation results based on the formula built into the contract. The carbon emission calculation formulas are pre-defined in the contract, and these formulas may be based on industry standards or specific calculation models.

[0036] The neural network model management contract is responsible for managing the storage address and version number associated with the model and storing this information in the blockchain;

[0037] The function of the early warning contract is to compare the calculated carbon emission index results with the preset index threshold, and decide whether to issue a notification to the enterprise and the regulator based on the comparison results.

[0038] S3. Each participating enterprise receives and stores its own activity data through its own business service system. Each business service system sends its own activity data to the blockchain, and calls the smart contract to successively reach consensus on the activity data storage, calculate the carbon emission index of the activity data to obtain the carbon emission index calculation result, store the carbon emission index calculation result in consensus and save it in the local database of each enterprise, and determine whether to issue an early warning notification based on the carbon emission index calculation result. If so, the early warning notification consensus is communicated to the regulator.

[0039] Figure 2 is a flowchart of data storage and calculation provided by an embodiment of the present invention, such as Figure 2 As shown, the S3 includes:

[0040] S31. Each participating enterprise party connects its own enterprise activity data to its own business service system through the REST API interface and saves it in its own enterprise local database;

[0041] In daily operations, enterprises will generate various activity data related to carbon emissions, such as various energy consumption generated in the production process. These data are an important basis for enterprises to monitor and calculate carbon emissions; in order to transmit these activity data to the business service system, enterprises connect the activity data to the business service system through the REST API interface.

[0042] When activity data is connected to the business service system through the REST API, it will first be saved in the local database within the enterprise; the purpose of saving activity data in the local database is to ensure that the enterprise has a complete record for subsequent processing, backup and quick access.

[0043] S32. Each business service system communicates with the blockchain through the blockchain SDK, uploads its own activity data to the blockchain, calls the data storage contract for storage, and distributes it to the regulator of the blockchain through the consensus mechanism;

[0044] Blockchain SDK is a development toolkit that provides an interface for interacting with the blockchain network. The business service system uses the API interface provided by the blockchain SDK to interact with the blockchain network, thereby uploading activity data to the blockchain.

[0045] By calling the data notarization contract, the activity data is written to the blockchain and verified by other nodes, and distributed to the regulators of the blockchain through the consensus mechanism.

[0046] S33. Each participating enterprise party calls the carbon index calculator contract to calculate the carbon emission index of its own activity data to obtain its own carbon emission index calculation results;

[0047] When the evidence consensus is completed, the carbon index calculator contract is called to pass the activity data that has just been successfully stored to the carbon index calculator contract. The carbon index calculator contract will process and calculate the activity data according to the built-in formula and logic to generate the carbon emission index calculation result.

[0048] It should be noted that the blockchain network may update the carbon index calculator contract regularly, especially when there are new calculation formulas and logics. By calling the carbon index calculator contract, the system can automatically use the latest version of the carbon index calculator (ie, the formula and logic) to ensure that the calculation results are based on the latest standards.

[0049] The entire process described above is highly automated. From the storage of activity data, consensus, calling of the carbon index calculator contract, to the generation of the final carbon emission index calculation results, all steps are completed without human intervention. This automated process improves the efficiency and accuracy of the system.

[0050] S34. Each participating enterprise party calls the data evidence contract to store the calculation results of their respective carbon emission indicators, distributes them to the regulator of the blockchain through the consensus mechanism, and saves the calculation results of their respective carbon emission indicators in the local database of their respective enterprises;

[0051] S35. Each participating enterprise party calls the early warning contract to compare the calculated results of its own carbon emission index with the preset index threshold. If the calculated results of the carbon emission index are greater than the preset index threshold, an early warning notification is issued, and the early warning notification consensus is communicated to the regulator based on the notification mechanism of the early warning contract. Otherwise, no early warning notification is issued.

[0052] After saving the carbon emission index calculation results, the system will automatically call the early warning contract to compare the carbon emission index calculation results with the preset index threshold. When the carbon emission index calculation results are less than or equal to the preset index threshold, no additional measures are required; when the carbon emission index calculation results are greater than the preset index threshold, it is necessary to activate the early warning contract notification mechanism to inform the regulator of the excessive carbon emissions. This notification mechanism can ensure that potential problems can be quickly noticed and handled.

[0053] S4. The regulator obtains the activity data and carbon emission index calculation results of all participating enterprises, and conducts model training and testing based on the long short-term memory neural network to obtain a carbon emission prediction model; saves the configuration file of the carbon emission prediction model to obtain the storage address, and calls the smart contract to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus;

[0054] Figure 3 is a flow chart of the model training provided by an embodiment of the present invention, such as Figure 3 As shown, the S4 includes:

[0055] S41. The regulator obtains the activity data and carbon emission index calculation results of all participating enterprises based on the data notarization contract;

[0056] S42. Classify the activity data and carbon emission index calculation results of all participating enterprises, and standardize the classified data;

[0057] The calculation results of corporate activity data and carbon emission indicators may include multiple dimensions, such as time series data, category data, continuous data, etc. In order to effectively train machine learning models, these data need to be classified and processed.

[0058] The classified data is standardized, and the standardization method includes:

[0059] (1) Maximum and minimum value normalization algorithm: The purpose is to scale the data range to [0, 1]. Normalization can eliminate the dimensional differences between data of different dimensions, so that the data can be processed at the same level, avoiding deviations caused by the large value range of some features during model training.

[0060] (2) Sliding average filtering algorithm: Sliding average is a commonly used method for processing time series data, which is used to smooth data and reduce noise and fluctuations. By calculating the average value of data within a certain window, sliding average can effectively eliminate short-term fluctuations and highlight long-term trends.

[0061] S43, dividing the classified data after standardization into a training set and a test set; iteratively training the training set based on the long short-term memory neural network to update the long short-term memory neural network, and predicting the test set through the long short-term memory neural network after the current iteration to obtain a prediction result; if the root mean square error between the prediction result and the true result is not within a preset range, the training set is iteratively trained for the next time based on the long short-term memory neural network after the current iteration until the root mean square error between the prediction result and the true result is within the preset range, stopping the iterative training, and using the long short-term memory neural network obtained by the last iteration as the carbon emission prediction model;

[0062] Long short-term memory (LSTM) neural network is a special recurrent neural network that can remember long-term dependencies and effectively prevent the gradient vanishing problem during training through its unique gating mechanism.

[0063] The regulator divides the standardized classified data into a training set and a test set. The training set is used to train the model, and the test set is used to evaluate the performance of the model. In a specific embodiment of the present invention, the data ratio of the training set and the test set is 9:1, which means that 90% of the data is used for training and 10% of the data is used for testing to evaluate the generalization ability of the model.

[0064] During the training and testing process, the long short-term memory neural network LSTM continuously adjusts parameters to make its prediction results as close to the true value as possible.

[0065] For example: the training set is iteratively trained for the first time based on the long short-term memory neural network, that is, the internal parameters of the long short-term memory neural network are updated to obtain the first updated long short-term memory neural network, and the test set is predicted by the first updated long short-term memory neural network to obtain the first prediction result; if the root mean square error RMSE between the first prediction result and the true result is not within the preset range; then the training set is iteratively trained for the second time based on the first updated long short-term memory neural network, that is, the internal parameters of the first updated long short-term memory neural network are updated to obtain the second updated long short-term memory neural network, and the test set is predicted by the second updated long short-term memory neural network to obtain the second prediction result; if the root mean square error RMSE between the second prediction result and the true result is also not within the preset range, then the model is trained again, and when it reaches 100 iterations, the test set is predicted by the 100th updated long short-term memory neural network to obtain the 100th prediction result; if the root mean square error RMSE between the 100th prediction result and the true result is within the preset range, it means that the training is qualified, and the 100th updated long short-term memory neural network is used as the carbon emission prediction model.

[0066] In a specific embodiment of the present invention, the preset range is less than 0.1. It should be noted that the present invention does not limit the preset range.

[0067] S44. Save the configuration file of the carbon emission prediction model to the cloud storage service to obtain the storage address, and call the neural network model management contract to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus.

[0068] Save the configuration file of the carbon emission prediction model (such as the model's weights, structure, hyperparameters, etc.) to the cloud storage service OSS to obtain the storage address (i.e., the URL in OSS);

[0069] When calling the neural network model management contract, the system will pass the storage address and version number of the carbon emission prediction model and other information as parameters to the neural network model management contract. After receiving these parameters, the neural network model management contract will write the storage address and version number of the carbon emission prediction model into the blockchain. Because blockchain data is tamper-proof and transparent, this ensures that all participating corporate parties can access the same model information (including the model's version number and storage address) and can verify the authenticity and integrity of the model.

[0070] S5. Each participating enterprise saves the storage address and version number of the latest carbon emission prediction model to the local database, so that each participating enterprise can predict the future carbon emissions of each participating enterprise based on the local activity data and carbon emission index calculation results based on the latest carbon emission prediction model.

[0071] When the carbon emission prediction model is updated, the configuration file of the updated carbon emission prediction model is saved to the cloud storage service to obtain the storage address, and the neural network model management contract is called to upload the storage address and version number of the updated carbon emission prediction model to the chain for evidence consensus.

[0072] That is, when the neural network model management contract updates the model information (such as the new version number and storage address), the updated model information will be broadcast to all nodes through the blockchain network.

[0073] Figure 4 is a flow chart of future carbon emission prediction provided by an embodiment of the present invention, such as Figure 4 As shown, the S5 includes:

[0074] S51. Each participating enterprise saves the storage address and version number of the latest carbon emission prediction model to a local database;

[0075] Although blockchain data is tamper-proof and open and transparent, the query efficiency of blockchain is relatively low, especially when faced with large amounts of data. Therefore, each participating enterprise node usually saves important information (such as the latest model information) to a local database for quick access and use.

[0076] S52. Each participating enterprise shall classify the activity data and carbon emission index calculation results in the local database and standardize the classified data;

[0077] Among them, the standardization processing method includes: maximum and minimum value normalization algorithm and sliding average filtering algorithm.

[0078] S53. Use the latest carbon emission prediction model to predict the future carbon emissions of each participating enterprise using the standardized classification data.

[0079] When the latest carbon emission prediction model is needed for prediction or analysis, each participating enterprise can query the latest version number and storage address of the carbon emission prediction model from the local database, and then download the corresponding model file from OSS for loading;

[0080] The standardized classified data is used to predict the future carbon emissions of each participating enterprise party through the latest carbon emission prediction model loaded. In a specific embodiment of the present invention, the carbon emissions of each participating enterprise party in the next six months can be predicted, and the prediction results can be saved in the local database.

[0081] Figure 5 is a structural diagram of a carbon emission monitoring system based on blockchain and neural network provided by an embodiment of the present invention; Figure 5 As shown, the system includes:

[0082] Node and system deployment unit 201, used for each participating enterprise and regulator to deploy blockchain nodes and business service systems;

[0083] A smart contract deployment unit 202, used to deploy a smart contract on a blockchain;

[0084] The smart contracts include: data storage contract, carbon index calculator contract, neural network model management contract, and early warning contract.

[0085] The evidence storage and calculation unit 203 is used for each participating enterprise to receive and store the activity data of each enterprise through its own business service system; each business service system sends its own activity data to the blockchain, and calls the smart contract to sequentially perform the activity data evidence storage consensus, calculate the carbon emission index of the activity data to obtain the carbon emission index calculation result, and store the carbon emission index calculation result in the local database of each enterprise, and determine whether to issue an early warning notification based on the carbon emission index calculation result. If so, the early warning notification consensus is sent to the regulator;

[0086] Figure 6 is a schematic diagram of the structure of the evidence storage and calculation unit provided in an embodiment of the present invention, such as Figure 6 As shown, the evidence storage and calculation unit 203 includes:

[0087] The receiving storage subunit 2031 is used for each participating enterprise party to access the activity data of their respective enterprises to their respective business service systems through the REST API interface and store the data in the local database of their respective enterprises;

[0088] Activity data evidence consensus subunit 2032, used for each business service system to communicate with the blockchain through the blockchain SDK, upload their respective activity data to the blockchain, call the data evidence contract for evidence, and distribute it to the regulator of the blockchain through the consensus mechanism;

[0089] The calculation subunit 2033 is used for each participating enterprise to call the carbon index calculator contract to calculate the carbon emission index of their respective activity data to obtain their respective carbon emission index calculation results;

[0090] The carbon emission index calculation result storage subunit 2034 is used for each participating enterprise to call the data evidence contract to store their respective carbon emission index calculation results, distribute them to the regulator of the blockchain through the consensus mechanism, and save their respective carbon emission index calculation results in the local database of each enterprise;

[0091] The early warning subunit 2035 is used for each participating enterprise to call the early warning contract to compare the calculated results of their respective carbon emission indicators with the preset indicator threshold. If the calculated results of the carbon emission indicators are greater than the preset indicator threshold, an early warning notification is issued, and the early warning notification consensus is communicated to the regulator based on the notification mechanism of the early warning contract. Otherwise, no early warning notification is issued.

[0092] The model training unit 204 is used for the regulator to obtain the activity data and carbon emission index calculation results of all participating enterprises, and to train and test the model based on the long short-term memory neural network to obtain the carbon emission prediction model; save the configuration file of the carbon emission prediction model to obtain the storage address, and call the smart contract to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus;

[0093] Figure 7 is a schematic diagram of the structure of the model training unit provided in an embodiment of the present invention, such as Figure 7 As shown, the model training unit 204 includes:

[0094] Acquisition subunit 2041 is used for the regulator to obtain the activity data and carbon emission index calculation results of all participating enterprises based on the data evidence contract;

[0095] The first standardization subunit 2042 is used to classify the activity data and carbon emission index calculation results of all participating enterprises, and perform standardization processing on the classified data;

[0096] The model training subunit 2043 is used to divide the classified data after the standardization processing into a training set and a test set; iteratively train the training set based on the long short-term memory neural network to update the long short-term memory neural network, and predict the test set through the long short-term memory neural network after the current iteration to obtain a prediction result; if the root mean square error between the prediction result and the true result is not within a preset range, the training set is iteratively trained for the next time based on the long short-term memory neural network after the current iteration until the root mean square error between the prediction result and the true result is within a preset range, the iterative training is stopped, and the long short-term memory neural network obtained by the last iteration is used as the carbon emission prediction model;

[0097] The model information evidence consensus subunit 2044 is used to save the configuration file of the carbon emission prediction model to the cloud storage service to obtain the storage address, and call the neural network model management contract to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus.

[0098] Prediction unit 205 is used for each participating enterprise to save the storage address and version number of the latest carbon emission prediction model to the local database, so that each participating enterprise can predict the future carbon emissions of each participating enterprise based on the local activity data and carbon emission index calculation results and the latest carbon emission prediction model.

[0099] When the carbon emission prediction model is updated, the configuration file of the updated carbon emission prediction model is saved to the cloud storage service to obtain the storage address, and the neural network model management contract is called to upload the storage address and version number of the updated carbon emission prediction model to the chain for evidence consensus.

[0100] Figure 8 is a schematic diagram of the structure of the prediction unit provided by an embodiment of the present invention, such as Figure 8 As shown, the prediction unit 205 includes:

[0101] The model information saving subunit 2051 is used for each participating enterprise to save the storage address and version number of the latest carbon emission prediction model into a local database;

[0102] The second standardization subunit 2052 is used for each participating enterprise to classify the activity data and carbon emission index calculation results in the local database and perform standardization on the classified data;

[0103] The prediction subunit 2053 is used to predict the future carbon emissions of each participating enterprise party using the latest carbon emission prediction model based on the standardized classification data.

[0104] Beneficial effects of the present invention:

[0105] The method of the present invention uses blockchain technology and neural network models to realize the automatic monitoring and processing of carbon emission data. The entire process from data access, evidence storage, calculation to early warning notification and model training can be automatically executed, reducing the need for manual operation and greatly improving the automation level of the system;

[0106] Through the carbon index calculator contract, the system can automatically calculate accurate carbon emission indicators based on the latest calculation formulas and logic. At the same time, through the classification and standardization of data, and the use of LSTM neural networks to train and predict historical data, the prediction accuracy of future carbon emission trends is improved;

[0107] The blockchain technology and distributed node structure adopted by the method of the present invention make the system highly scalable and flexible. Whether the number of participating enterprises increases or the complexity of the model increases, the system can adapt by adding nodes or adjusting parameters to maintain good performance.

[0108] Activity data and carbon emission index calculation results are stored through blockchain technology. The immutability and distributed consensus mechanism of blockchain are utilized to ensure the authenticity and integrity of the data. All data on the chain are verified by consensus, avoiding the possibility of data tampering and enhancing the security of the system.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A carbon emission monitoring method based on blockchain and neural network, characterized in that: include: S1. Each participating enterprise and regulator deploys blockchain nodes and business service systems; S2. Deploy smart contracts on the blockchain; S3. Each participating enterprise receives and stores its own activity data through its own business service system. Each business service system sends its own activity data to the blockchain, and calls the smart contract to successively reach consensus on the activity data storage, calculate the carbon emission index of the activity data to obtain the carbon emission index calculation result, store the carbon emission index calculation result in consensus and save it in the local database of each enterprise, and determine whether to issue an early warning notification based on the carbon emission index calculation result. If so, the early warning notification consensus is communicated to the regulator. S4. The regulator obtains the activity data and carbon emission index calculation results of all participating enterprises, and conducts model training and testing based on the long short-term memory neural network to obtain a carbon emission prediction model; saves the configuration file of the carbon emission prediction model to obtain the storage address, and calls the smart contract to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus; S5. Each participating enterprise saves the storage address and version number of the latest carbon emission prediction model to the local database, so that each participating enterprise can predict the future carbon emissions of each participating enterprise based on the local activity data and carbon emission index calculation results based on the latest carbon emission prediction model.

2. The method according to claim 1, characterized in that: The smart contracts include: data storage contract, carbon index calculator contract, neural network model management contract, and early warning contract.

3. The method according to claim 2, characterized in that The S3 includes: Each participating enterprise connects its own enterprise activity data to its own business service system through the REST API interface and saves it in its own enterprise local database; Each business service system communicates with the blockchain through the blockchain SDK, uploads its own activity data to the blockchain, calls the data storage contract for storage, and distributes it to the blockchain regulator through the consensus mechanism; Each participating enterprise party calls the carbon index calculator contract to calculate the carbon emission index of its own activity data to obtain its own carbon emission index calculation results; Each participating enterprise calls the data notarization contract to store the calculation results of their respective carbon emission indicators, distributes them to the regulator of the blockchain through the consensus mechanism, and saves the calculation results of their respective carbon emission indicators in their respective local databases; Each participating enterprise calls the early warning contract to compare the calculated results of its own carbon emission index with the preset index threshold. If the calculated result of the carbon emission index is greater than the preset index threshold, an early warning notification will be issued, and the early warning notification consensus will be communicated to the regulator based on the notification mechanism of the early warning contract. Otherwise, no early warning notification will be issued.

4. The method according to claim 2, characterized in that: The S4 includes: The regulator obtains the activity data and carbon emission index calculation results of all participating companies based on the data notarization contract; Classify the activity data and carbon emission index calculation results of all participating enterprises, and standardize the classified data; The classified data after standardization is divided into a training set and a test set; the training set is iteratively trained based on the long short-term memory neural network to update the long short-term memory neural network, and the test set is predicted by the long short-term memory neural network after the current iteration to obtain the prediction result; if the root mean square error between the prediction result and the true result is not within the preset range, the training set is iteratively trained for the next time based on the long short-term memory neural network after the current iteration until the root mean square error between the prediction result and the true result is within the preset range, the iterative training is stopped, and the long short-term memory neural network obtained by the last iteration is used as the carbon emission prediction model; Save the configuration file of the carbon emission prediction model to the cloud storage service to obtain the storage address, and call the neural network model management contract to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus.

5. The method according to claim 4, characterized in that: When the carbon emission prediction model is updated, the configuration file of the updated carbon emission prediction model is saved to the cloud storage service to obtain the storage address, and the neural network model management contract is called to upload the storage address and version number of the updated carbon emission prediction model to the chain for evidence consensus.

6. The method according to claim 5, characterized in that The S5 includes: Each participating enterprise saves the storage address and version number of the latest carbon emission prediction model to the local database; Each participating enterprise will classify the activity data and carbon emission index calculation results in the local database and standardize the classified data; The standardized classified data is used through the latest carbon emission prediction model to predict the future carbon emissions of each participating enterprise.

7. A carbon emission monitoring system based on blockchain and neural network, characterized in that: include: Node and system deployment unit, used by each participating enterprise and regulator to deploy blockchain nodes and business service systems; Smart contract deployment unit, used to deploy smart contracts on the blockchain; The evidence storage and calculation unit is used for each participating enterprise to receive and store the activity data of each enterprise through its own business service system; each business service system sends its own activity data to the blockchain, and calls the smart contract to successively reach consensus on the activity data evidence storage, calculate the carbon emission index of the activity data to obtain the carbon emission index calculation result, and store the carbon emission index calculation result in the local database of each enterprise, and determine whether to issue an early warning notification based on the carbon emission index calculation result. If so, the early warning notification consensus is communicated to the regulator; The model training unit is used by the regulator to obtain the activity data and carbon emission index calculation results of all participating enterprises, and to train and test the model based on the long short-term memory neural network to obtain the carbon emission prediction model; save the configuration file of the carbon emission prediction model to obtain the storage address, and call the smart contract to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus; The prediction unit is used for each participating enterprise to save the storage address and version number of the latest carbon emission prediction model to the local database, so that each participating enterprise can predict the future carbon emissions of each participating enterprise based on the local activity data and carbon emission index calculation results and the latest carbon emission prediction model.

8. The system according to claim 7, characterized in that: The smart contracts include: data storage contract, carbon index calculator contract, neural network model management contract, and early warning contract.

9. The system according to claim 8, characterized in that The evidence storage and calculation unit includes: A receiving storage subunit is used for each participating enterprise to connect the activity data of each enterprise to their respective business service system through the REST API interface and save it in the local database of each enterprise; The activity data evidence consensus subunit is used for each business service system to communicate with the blockchain through the blockchain SDK, upload their respective activity data to the blockchain, call the data evidence contract for evidence storage, and distribute it to the regulator of the blockchain through the consensus mechanism; The calculation subunit is used for each participating enterprise to call the carbon index calculator contract to calculate the carbon emission index of their respective activity data to obtain their respective carbon emission index calculation results; The carbon emission index calculation result storage subunit is used for each participating enterprise to call the data evidence contract to store their respective carbon emission index calculation results, distribute them to the regulator of the blockchain through the consensus mechanism, and save their respective carbon emission index calculation results in their respective enterprise's local databases; The early warning sub-unit is used for each participating enterprise to call the early warning contract to compare the calculated results of their respective carbon emission indicators with the preset indicator threshold. If the calculated results of the carbon emission indicators are greater than the preset indicator threshold, an early warning notification will be issued, and the early warning notification consensus will be communicated to the regulator based on the notification mechanism of the early warning contract. Otherwise, no early warning notification will be issued.

10. The system according to claim 8, characterized in that The model training unit comprises: The acquisition subunit is used by the regulator to obtain the activity data and carbon emission index calculation results of all participating enterprises based on the data notarization contract; The first standardization sub-unit is used to classify the activity data and carbon emission index calculation results of all participating enterprises and standardize the classified data; The model training subunit is used to divide the classified data after standardization into a training set and a test set; iteratively train the training set based on the long short-term memory neural network to update the long short-term memory neural network, and predict the test set through the long short-term memory neural network after the current iteration to obtain the prediction result; if the root mean square error between the prediction result and the true result is not within a preset range, the training set is iteratively trained for the next time based on the long short-term memory neural network after the current iteration until the root mean square error between the prediction result and the true result is within the preset range, the iterative training is stopped, and the long short-term memory neural network obtained by the last iteration is used as the carbon emission prediction model; The model information evidence consensus sub-unit is used to save the configuration file of the carbon emission prediction model to the cloud storage service to obtain the storage address, and call the neural network model management contract to upload the storage address and version number of the carbon emission prediction model to the chain for evidence consensus.

Images