WO2024055546A1 - 动力电池回收量预测方法、装置及介质 - Google Patents

动力电池回收量预测方法、装置及介质 Download PDF

Info

Publication number
WO2024055546A1
WO2024055546A1 PCT/CN2023/081916 CN2023081916W WO2024055546A1 WO 2024055546 A1 WO2024055546 A1 WO 2024055546A1 CN 2023081916 W CN2023081916 W CN 2023081916W WO 2024055546 A1 WO2024055546 A1 WO 2024055546A1
Authority
WO
WIPO (PCT)
Prior art keywords
power battery
battery recycling
prediction model
training
carbon emission
Prior art date
Application number
PCT/CN2023/081916
Other languages
English (en)
French (fr)
Inventor
余海军
李爱霞
谢英豪
李长东
Original Assignee
广东邦普循环科技有限公司
湖南邦普循环科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 广东邦普循环科技有限公司, 湖南邦普循环科技有限公司 filed Critical 广东邦普循环科技有限公司
Publication of WO2024055546A1 publication Critical patent/WO2024055546A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services

Definitions

  • This application relates to the technical field of carbon emissions, for example, to a method, device and medium for predicting the recycling amount of a power battery.
  • This application provides a method, device and medium for predicting the recycling amount of power batteries to solve the technical problems of low accuracy and complex prediction of the recycling amount of power batteries in related technologies.
  • Embodiments of the present application provide a method for predicting the recycling amount of a power battery, which includes: obtaining the carbon emission data of the power battery; inputting the carbon emission data into a power battery recycling prediction model, and outputting the recycling amount of the power battery; wherein, constructing the The power battery recycling prediction model includes: obtaining the historical carbon emission data of the power battery and the power battery recycling amount corresponding to the historical carbon emission data, and combining the historical carbon emission data and the power corresponding to the historical carbon emission data.
  • the battery recycling amount is used as the first training data; based on the first training data, the constructed initial power battery recycling prediction model is trained, and the trained power battery recycling prediction model is output.
  • training the constructed initial power battery recycling prediction model according to the first training data, thereby outputting the trained power battery recycling prediction model including: constructing an initial power battery recycling prediction model; according to An optimization algorithm is used to initialize the weights and thresholds of the initial power battery recycling prediction model; and based on the first training data, the initial power battery recovery prediction model is The power battery recycling prediction model is iteratively trained, and the power battery recycling prediction model after training is output.
  • iteratively training the initial power battery recycling prediction model based on the first training data, and outputting the power battery recycling prediction model after training is completed includes: training the initial power battery recycling prediction model Iterative training is performed so that in each training process, the output value and the first training data are calculated based on the historical carbon emission data in the first training data input to the initial power battery recycling prediction model.
  • the error index function of the power battery recycling amount in and calculate and update the weights and thresholds of the initial power battery recycling prediction model according to the output value and the error index function until the error index function is less than the preset allowable value, output the power battery recycling prediction model after training.
  • the method further includes: recording the carbon emission data and the power battery recycling amount, and converting the power battery recycling amount to The carbon emission data and the power battery recycling amount are used as the second training data; according to the second training data, the power battery recycling prediction model is retrained, and the retrained power battery recycling prediction model is output.
  • the carbon emission data of the power battery includes the carbon emission data of the power battery in the production stage, transportation stage and use stage;
  • the historical carbon emission data of the power battery includes the power battery in the production stage, Historical carbon emissions data during transportation and use phases.
  • This application also provides a power battery recycling amount prediction device, including: an initial data module, a prediction module and a modeling module.
  • the initial data module is configured to obtain the carbon emission data of the power battery;
  • the prediction module is configured to input the carbon emission data into the power battery recycling prediction model and output the power battery recycling amount;
  • the modeling module It includes: a first training data sub-module and a training sub-module;
  • the first training data sub-module is configured to obtain the historical carbon emission data of the power battery and the power battery recycling amount corresponding to the historical carbon emission data, and will The historical carbon emission data and the power battery recycling amount corresponding to the historical carbon emission data are used as the first training data;
  • the training sub-module is configured to calculate the initial power battery recycling prediction model based on the first training data. Carry out training and output the trained power battery recycling prediction model.
  • the training sub-module is configured to train the constructed initial power battery recycling prediction model based on the first training data in the following manner, and output the trained power battery recycling prediction model, including: constructing an initial power battery recycling prediction model a power battery recycling prediction model; initialize the weights and thresholds of the initial power battery recycling prediction model according to an optimization algorithm; perform iterative training on the initial power battery recycling prediction model according to the first training data, Output the power battery recycling prediction model after training.
  • iteratively training the initial power battery recycling prediction model based on the first training data, and outputting the power battery recycling prediction model after training is completed includes: training the initial power battery recycling prediction model Iterative training is performed so that in each training process, the output value and the first training data are calculated based on the historical carbon emission data in the first training data input to the initial power battery recycling prediction model.
  • the error index function of the power battery recycling amount in and calculate and update the weights and thresholds of the initial power battery recycling prediction model according to the output value and the error index function until the error index function is less than the preset allowable value, output the power battery recycling prediction model after training.
  • the embodiment of the present application also includes: a retraining module.
  • the retraining module is configured to record the carbon emission data and the power battery recycling amount, and use the carbon emission data and the power battery recycling amount as the second training data; according to the second training number, The power battery recycling prediction model is retrained, and the retrained power battery recycling prediction model is output.
  • the carbon emission data of the power battery includes the carbon emission data of the power battery in the production stage, transportation stage and use stage;
  • the historical carbon emission data of the power battery includes the power battery in the production stage, Historical carbon emissions data during transportation and use phases.
  • This application also provides a computer-readable storage medium, including a stored computer program; wherein, when running, the computer program controls the device where the computer-readable storage medium is located to perform the power battery recycling amount as described in any one of the above. method of prediction.
  • Figure 1 A flow chart of a power battery recycling amount prediction method provided by an embodiment of the present application
  • Figure 2 A flow chart of retraining in a power battery recycling amount prediction method provided by an embodiment of the present application
  • Figure 3 A flow chart for constructing a power battery recycling prediction model in a power battery recycling amount prediction method provided by the embodiment of the present application
  • Figure 4 Structural diagram of the initial power battery recycling prediction model in a power battery recycling amount prediction method provided by the embodiment of the present application;
  • Figure 5 A schematic structural diagram of a power battery recycling amount prediction device based on carbon emissions provided by an embodiment of the present application.
  • Figure 1 is a method for predicting the recycling amount of a power battery provided by an embodiment of the present application, including the following steps S101-S102.
  • the carbon emission data of the power battery includes the carbon emission data of the power battery in the production stage, transportation stage and use stage.
  • the carbon emission data of the power battery in the production stage, transportation stage and use stage can be obtained in the form of a sequence; for example, the carbon emission data generated in the production stage is taken as X1, Let the carbon emission data generated during the transportation phase be regarded as X2, and the carbon emission data generated during the use phase be regarded as Input data for the battery recycling prediction model.
  • the carbon emissions in the production stage include carbon emissions from raw material acquisition and battery assembly.
  • the raw material acquisition stage it is first necessary to collect relevant data on the raw materials required for manufacturing batteries, and determine the types of raw materials required for the corresponding type of energy storage battery in this embodiment and their requirements through the single product specification and bill of materials corresponding to this embodiment.
  • raw materials include but are not limited to positive electrode materials, negative electrode materials, separators, electrolytes, battery packs, etc.
  • the production line data of each raw material analyzing the carbon emissions produced by producing the material at a certain quantitative output, the unit mass of the material can be obtained.
  • the carbon emissions generated are calculated as the carbon emissions during the raw material acquisition phase of the material.
  • the battery assembly stage it mainly includes battery manufacturing, circuit board manufacturing and battery pack assembly; obtain the carbon emissions caused by the process and energy consumption during the battery manufacturing process, and calculate the production unit mass during the battery assembly stage
  • the carbon emissions of the battery are calculated, and the carbon emissions of all raw materials required for manufacturing and the manufacturing carbon emissions are summed to obtain the carbon emissions corresponding to the production stage of the power battery of this embodiment.
  • the X3 acquisition principle of carbon emission data generated during the use phase is as follows: the power loss of power batteries during the operating phase is also one of the factors that affect its carbon emissions throughout its life cycle. Power batteries will generate electrical energy during the operating phase due to the charging and discharging efficiency. This part of the loss is Electric energy should be calculated into the carbon emissions during the operation phase of the energy storage battery. Different power generation types (such as thermal power, hydropower, nuclear power, etc.) need to be determined based on the local power generation energy structure. The percentage of the power generation structure and the carbon emission intensity of this energy generation are used to obtain the carbon emissions of the loss of this part of the electricity.
  • S102 Input the carbon emission data into the power battery recycling prediction model, and output the power battery recycling amount.
  • each sequence of carbon emission data corresponds to a power battery recycling amount for output.
  • steps S103-S104 are also included:
  • S103 Record the carbon emission data and the power battery recycling amount, and use the carbon emission data and the power battery recycling amount as the second training data.
  • S104 Retrain the power battery recycling prediction model according to the second training data, and output the retrained power battery recycling prediction model.
  • the power battery recycling prediction model is retrained, so that the power battery recycling
  • the recycling prediction model can be automatically updated and revised to ensure the accuracy and accuracy of the power battery recycling prediction model, so as to improve the accuracy of power battery recycling volume prediction.
  • the construction steps of the power battery recycling prediction model include the following steps S201-S202:
  • S201 Obtain the historical carbon emission data of the power battery and the power battery recycling quantity corresponding to the historical carbon emission data, and use the historical carbon emission data and the power battery recycling quantity corresponding to the historical carbon emission data as the first training data.
  • the historical carbon emission data of the power battery includes the historical carbon emission data of the power battery in the production stage, transportation stage and use stage.
  • Y3 is the historical carbon emission data generated during the power battery emission phase.
  • the recycling amount of power batteries corresponding to historical carbon emission data is based on this carbon emission data.
  • 70% of the data in the first training data is selected as training data, 15% of the data is used as verification data, and 15% of the data is used as test data.
  • S202 Train the constructed initial power battery recycling prediction model according to the first training data, and output the trained power battery recycling prediction model.
  • the initial power battery recycling prediction model constructed is an error back propagation (BP) neural network model, including: an input layer, a hidden layer and output layer; among them, the input layer includes three nodes corresponding to the production stage, transportation stage and usage stage, the output layer includes one node corresponding to the number of power battery recycling, and the number of hidden layer nodes is set according to the actual demand. .
  • BP error back propagation
  • training the constructed initial power battery recycling prediction model based on the first training data, and outputting the trained power battery recycling prediction model includes: constructing an initial power battery recycling prediction model Prediction model; according to the optimization algorithm, initialize the weights and thresholds of the initial power battery recycling prediction model; perform iterative training on the initial power battery recycling prediction model according to the first training data, and the output training is completed The final power battery recycling prediction model.
  • the optimization algorithm in the embodiment of the present application is the Levenberg-Marquardt algorithm, which is usually used to minimize the objective function of the nonlinear least squares method.
  • the weights and thresholds of the initial power battery recycling prediction model are initialized according to the optimization algorithm to ensure that the pre-training network model can meet the requirements of subsequent iterative training.
  • Model accuracy requirements and iteratively train the initial power battery recycling prediction model through the first training data, ensuring that the trained power battery recycling prediction model has high precision and accuracy, and making the trained power battery recycling prediction
  • the model avoids the interference of human factors, making the model more reasonable and objective.
  • iteratively training the initial power battery recycling prediction model based on the first training data, and outputting the power battery recycling prediction model after training is completed includes:
  • the initial power battery recycling prediction model is iteratively trained, so that in each training process, the historical data in the first training data input to the initial power battery recycling prediction model will be Historical carbon emission data, calculate the error index function between the output value and the power battery recycling amount in the first training data, and calculate and update the weight sum of the initial power battery recycling prediction model based on the output value and the error index function Threshold until the error index function is less than the preset allowed value, the power battery recycling prediction model after training is output.
  • the model output and error index function E(w) are calculated, Among them, Y i is the expected model output vector; Y′ i is the actual model output vector; P is the number of samples; w is the vector composed of model weights and thresholds; e i (w) is the error.
  • the test set data and verification set data it is also necessary to input the test set data and verification set data to fit the residual error and mean square error of the power battery recycling prediction model. Calculations are performed to ensure that the power battery recycling prediction model meets error requirements, has a high degree of fitting and a high degree of linearization.
  • the mean square error is the average squared difference between the output and the target output, and the lower the value, the better, i.e. zero means no error.
  • the power battery recycling prediction model can have high-precision and high-accuracy prediction capabilities, and at the same time, prediction through the network model avoids the existing manual prediction.
  • High complexity, and in each training process, the weights and thresholds of the network model are updated through the error index function, so that the initial power battery recycling prediction model gradually converges, so that when the error index function is less than the preset allowed value, This ensures that the power battery recycling prediction model after training can be output accurately and stably.
  • the established initial power battery feedback is collected by collecting the historical carbon emission data of the power battery and using the number of power battery recycling corresponding to the historical carbon emission data as training data.
  • the recycling prediction model is trained, so that the power battery recycling prediction model can be accurately obtained, so that the carbon emission data of the power battery can be directly input in the future, and the recycling amount of the power battery can be accurately output, which improves the company staff's understanding of the power battery.
  • the efficiency of formulating battery recycling treatment plans avoids the high complexity of existing manual subjective predictions, assists decision-making on battery recycling amounts, and improves the accuracy and rationality of power battery recycling.
  • this application also provides a power battery recycling amount prediction device, including: an initial data module 301, a prediction module 302 and a modeling module 303.
  • the initial data module 301 is configured to obtain carbon emission data of the power battery.
  • the prediction module 302 is configured to input the carbon emission data into a power battery recycling prediction model and output the power battery recycling amount.
  • the modeling module 303 includes: a first training data sub-module 3031 and a training sub-module 3032.
  • the first training data sub-module 3031 is configured to obtain the historical carbon emission data of the power battery and the power battery recycling amount corresponding to the historical carbon emission data, and combine the historical carbon emission data and the corresponding data of the historical carbon emission data.
  • the amount of power battery recycling is used as the first training data.
  • the training sub-module 3032 is configured to train the constructed initial power battery recycling prediction model based on the first training data, and output the trained power battery recycling prediction model.
  • the training sub-module 3032 is configured to train the constructed initial power battery recycling prediction model based on the first training data in the following manner, and output the trained power battery recycling prediction
  • the model includes: constructing an initial power battery recycling prediction model; performing initialization operations on weights and thresholds of the initial power battery recycling prediction model according to an optimization algorithm; and performing initialization operations on the initial power battery recycling prediction model based on the first training data.
  • the prediction model is iteratively trained and the power battery recycling prediction model after training is output.
  • iteratively training the initial power battery recycling prediction model based on the first training data, thereby outputting the power battery recycling prediction model after training includes:
  • the initial power battery recycling prediction model is trained iteratively, so that in each training process, an output value will be calculated based on the historical carbon emission data in the first training data input to the initial power battery recycling prediction model. and the error index function of the power battery recycling amount in the first training data, and based on the output value and the error index function, calculate and update the weights and thresholds of the initial power battery recycling prediction model network until the error index function is less than the predetermined When the allowed value is set, the power battery recycling prediction model after training is output.
  • this embodiment also includes: a retraining module 304.
  • the retraining module 304 is configured to record the carbon emission data and the power battery recycling amount, and use the carbon emission data and the power battery recycling amount as the second training data; according to the second training data, The power battery recycling prediction model is retrained, and the retrained power battery recycling prediction model is output.
  • the carbon emission data of the power battery includes the carbon emission data of the power battery in the production stage, transportation stage and use stage;
  • the historical carbon emission data of the power battery includes the carbon emission data of the power battery in the production stage. Historical carbon emissions data at stage, transport stage and use stage.
  • the embodiment of the present application collects the historical carbon emission data of the power battery and uses the power battery recycling amount corresponding to the historical carbon emission data as training data to train the initial power battery recycling prediction model established, so that the power battery can be accurately obtained
  • the recycling prediction model allows the carbon emission data of the power battery to be directly input in the future, and the recycling amount of the power battery can be accurately output, which improves the efficiency of the company staff in formulating the power battery recycling treatment plan and avoids the existing
  • the high complexity of manual and subjective predictions assists decision-making on the number of batteries to be recycled, which improves the accuracy and rationality of power battery recycling.
  • This application also provides a terminal device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor.
  • a terminal device including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor.
  • the processor executes the computer program, any one of the above is implemented.
  • the terminal device of this embodiment includes: a processor, a memory, and a computer program and computer instructions stored in the memory and executable on the processor.
  • the processor executes the computer program, it implements multiple steps in the above-mentioned Embodiment 1, such as steps S101 to S102 shown in FIG. 1 .
  • the processor executes the computer program, it implements the functions of multiple modules/units in the above device embodiment, such as the prediction module 302.
  • the computer program may be divided into one or more modules/units, and the one or more modules/units are stored in the memory and executed by the processor to complete the present application.
  • the one or more modules/units may be a series of computer program instruction segments capable of completing specific functions.
  • the instruction segments are used to describe the execution process of the computer program in the terminal device.
  • the prediction module 302 is configured to input the carbon emission data into a power battery recycling prediction model and output the power battery recycling amount.
  • the terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc.
  • the terminal device may include, but is not limited to, a processor and a memory.
  • a processor and a memory.
  • the schematic diagram is only an example of the terminal equipment and does not constitute a limitation of the terminal equipment. It may include more or less components than shown in the figure, or combine some components, or different components.
  • the terminal equipment also Can include input and output devices, network access devices, buses, etc.
  • the so-called processor can be a central processing unit (Central Processing Unit, CPU), or other general-purpose processor, digital signal processor (Digital Signal Processor, DSP), application specific integrated circuit (Application Specific Integrated Circuit, ASIC), off-the-shelf Programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.
  • the general processor may be a microprocessor or the processor may be any conventional processor, etc.
  • the processor is the control center of the terminal device and uses a variety of interfaces and lines to connect multiple parts of the entire terminal device.
  • the memory may be configured to store the computer program and/or module, and the processor implements the terminal device by running or executing the computer program and/or module stored in the memory, and calling data stored in the memory. of multiple functions.
  • the memory may mainly include a stored program area and a stored data area, wherein the stored program area may store an operating system, an application program required for at least one function, etc.; the stored data area may store data created according to the use of the mobile terminal, etc.
  • the memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD) card , Flash Card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.
  • non-volatile memory such as hard disk, memory, plug-in hard disk, smart memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD) card , Flash Card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.
  • modules/units integrated with the terminal device are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium.
  • the present application can implement all or part of the processes in the methods of the above embodiments, which can also be completed by instructing relevant hardware through a computer program.
  • the computer program can be stored in a computer-readable storage medium, and the computer can When the program is executed by the processor, the steps of the above multiple method embodiments can be implemented.
  • the computer program includes computer program code, which may be in the form of source code, object code, executable file or some intermediate form.
  • the computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (Read-Only Memory, ROM) , Random Access Memory (RAM), electrical carrier signals, telecommunications signals, and software distribution media, etc. It should be noted that the content contained in the computer-readable medium can be appropriately added or deleted according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not Including electrical carrier signals and telecommunications signals.
  • this application also provides a computer-readable storage medium.
  • the computer-readable storage medium The substance includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the power battery recycling amount prediction method as described in any one of the above embodiments.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • General Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • Health & Medical Sciences (AREA)
  • Marketing (AREA)
  • General Health & Medical Sciences (AREA)
  • Development Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Educational Administration (AREA)
  • Biophysics (AREA)
  • Operations Research (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Primary Health Care (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Quality & Reliability (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Secondary Cells (AREA)

Abstract

本申请公开了一种动力电池回收量预测方法及装置，方法包括：获取动力电池的碳排放数据；将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量；其中，构建所述动力电池回收预测模型，包括：获取所述动力电池的历史碳排放数据以及对应所述历史碳排放数据的动力电池回收量，将所述历史碳排放数据以及所述历史碳排放数据对应的动力电池回收量作为第一训练数据；根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型。

Description

动力电池回收量预测方法、装置及介质
本申请要求在2022年09月16日提交中国专利局、申请号为202211132754.3的中国专利申请的优先权，该申请的全部内容通过引用结合在本申请中。
技术领域
本申请涉及碳排放技术领域，例如涉及一种动力电池回收量预测方法、装置及介质。
背景技术
制造新电池需要大量能源，电池回收可节省更多能源，同时电池材料是可以循环的，发展动力电池梯次利用，做好动力电池的环保处理，充分挖掘动力电池的价值，能够在客观上减少碳排放。
但是在动力电池的实际回收过程中也会存在碳的排放，若过量的回收会导致碳排放不降反增，从而难以实现减少碳排放，而对动力电池回收数量进行预测主要是根据人为主观进行确定，难以得到一个准确且合适的动力电池回收数量，不利于人们对动力电池回收量进行准确的预测，使得预测复杂性高的同时精确度低。
发明内容
本申请提供了一种动力电池回收量预测方法、装置及介质，以解决相关技术中对动力电池回收量预测精确度低、预测复杂的技术问题。
本申请实施例提供了一种动力电池回收量预测方法，包括：获取动力电池的碳排放数据；将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量；其中，构建所述动力电池回收预测模型，包括：获取所述动力电池的历史碳排放数据以及对应所述历史碳排放数据的动力电池回收量，将所述历史碳排放数据以及所述历史碳排放数据对应的动力电池回收量作为第一训练数据；根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型。
作为可选方案，所述根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，从而输出得到训练后的动力电池回收预测模型，包括：构建初始动力电池回收预测模型；根据最优化算法，对所述初始动力电池回收预测模型的权值和阈值进行初始化操作；根据所述第一训练数据，对所述初始动 力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型。
作为可选方案，所述根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型，包括：对所述初始动力电池回收预测模型进行迭代训练，以使在每一次训练过程中，将根据输入至所述初始动力电池回收预测模型的所述第一训练数据中的历史碳排放数据，计算得到输出值与所述第一训练数据中的动力电池回收量的误差指标函数，并根据所述输出值和所述误差指标函数，来计算并更新初始动力电池回收预测模型的权值和阈值，直至所述误差指标函数小于预设允许值时，输出训练完成后的动力电池回收预测模型。
作为可选方案，在所述将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量之后，还包括：记录所述碳排放数据以及所述动力电池回收量，将所述碳排放数据以及所述动力电池回收量作为第二训练数据；根据所述第二训练数据，对所述动力电池回收预测模型进行再训练，输出再训练后的动力电池回收预测模型。
作为可选方案，所述动力电池的碳排放数据包括所述动力电池在生产阶段、运输阶段和使用阶段的碳排放数据；所述动力电池的历史碳排放数据包括所述动力电池在生产阶段、运输阶段和使用阶段的历史碳排放数据。
本申请还提供一种动力电池回收量预测装置，包括：初始数据模块、预测模块和建模模块。
所述初始数据模块，设置为获取动力电池的碳排放数据；所述预测模块，设置为将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量；所述建模模块，包括：第一训练数据子模块和训练子模块；所述第一训练数据子模块，设置为获取所述动力电池的历史碳排放数据以及对应所述历史碳排放数据的动力电池回收量，将将所述历史碳排放数据以及所述历史碳排放数据对应的动力电池回收量作为第一训练数据；所述训练子模块，设置为根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型。
作为可选方案，所述训练子模块设置为通过以下方式根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出得到训练后的动力电池回收预测模型，包括：构建初始动力电池回收预测模型；根据最优化算法，对所述初始动力电池回收预测模型的权值和阈值进行初始化操作；根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型。
作为可选方案，所述根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型，包括：对所述初始动力电池回收预测模型进行迭代训练，以使在每一次训练过程中，将根据输入至所述初始动力电池回收预测模型的所述第一训练数据中的历史碳排放数据，计算得到输出值与所述第一训练数据中的动力电池回收量的误差指标函数，并根据所述输出值和所述误差指标函数，来计算并更新初始动力电池回收预测模型的权值和阈值，直至所述误差指标函数小于预设允许值时，输出训练完成后的动力电池回收预测模型。
作为可选方案，本申请实施例还包括：再训练模块。
所述再训练模块，设置为记录所述碳排放数据以及所述动力电池回收量，将所述碳排放数据以及所述动力电池回收量作为第二训练数据；根据所述第二训练数，对所述动力电池回收预测模型进行再训练，输出再训练后的动力电池回收预测模型。
作为可选方案，所述动力电池的碳排放数据包括所述动力电池在生产阶段、运输阶段和使用阶段的碳排放数据；所述动力电池的历史碳排放数据包括所述动力电池在生产阶段、运输阶段和使用阶段的历史碳排放数据。
本申请还提供一种计算机可读存储介质，包括存储的计算机程序；其中，所述计算机程序在运行时控制所述计算机可读存储介质所在的设备执行如上任一项所述的动力电池回收量预测方法。
附图说明
图1：为本申请实施例所提供的一种动力电池回收量预测方法的流程图；
图2：为本申请实施例所提供的一种动力电池回收量预测方法中再训练的流程图；
图3：为本申请实施例所提供的一种动力电池回收量预测方法中构建动力电池回收预测模型的流程图；
图4：为本申请实施例所提供的一种动力电池回收量预测方法中初始动力电池回收预测模型的结构图；
图5：为本申请实施例所提供的一种基于碳排放的动力电池回收量预测装置的结构示意图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行描 述，显然，所描述的实施例仅仅是本申请一部分实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
实施例一
请参照图1，为本申请实施例提供的一种动力电池回收量预测方法，包括以下步骤S101-S102。
S101：获取动力电池的碳排放数据。
所述动力电池的碳排放数据包括动力电池在生产阶段、运输阶段和使用阶段的碳排放数据。
需要说明的是，在本实施例中，动力电池在生产阶段、运输阶段和使用阶段的碳排放数据可以通过序列的形式进行获取；示例性地，将生产阶段所产生的碳排放数据作为X1，将运输阶段所产生的碳排放数据作为X2，将使用阶段所产生的碳排放数据作为X3，从而作为一个动力电池碳排放数据的序列A＝[X1,X2,X3]，并将序列A作为动力电池回收预测模型的输入数据。
生产阶段所产生的碳排放数据X1获取原理为：生产阶段碳排放包括原料获取和电池组装的碳排放。对于原料获取阶段，首先需要收集制造电池所需原材料的相关数据，通过本实施例对应的单一产品规格书和物料清单来确定本实施例对应类型的储能电池所需原材料的种类以及其所需质量，原材料包括但不限于正极材料、负极材料、隔膜、电解液、电池包等。进而需要确定获取每类原材料所产生的碳排放，通过获取每种原材料的生产线数据，分析该种材料在一个定量产量下，生产该种材料所产生的碳排放量，得到获得该种材料单位质量所产生的碳排放，计算得出该材料在原材料获取阶段的碳排放。对于电池组装阶段，主要包括电池的制造、电路板的制造和电池包的组装；获取电池在制造过程中，因所用工艺和消耗能源产生的碳排放量，计算得出在电池组装阶段生产单位质量电池的碳排放，并将制造所需的所有原材料碳排放和制造碳排放求和，从而得到对应于本实施例的动力电池在生产阶段的碳排放。
运输阶段所产生的碳排放数据X2获取原理为：运输阶段需收集动力电池出厂运输时，使用不同运输工具完成运输所消耗能源的量，以及该种能源的碳排放，基于上述数据确定该阶段的碳排放量。
使用阶段所产生的碳排放数据X3获取原理为：动力电池运行阶段的电能损耗也是影响其全生命周期碳排放的因素之一，动力电池因充放电效率在运行阶段会产生电能，这部分损耗的电能应计算到储能电池运行阶段碳排放中，其中需要基于当地发电能源结构确定不同发电类型(例如：火电、水电、核电等)所 占发电结构的百分比，以及该种能源发电碳排放强度，得到损耗这部分电能的碳排放量。
S102：将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量。
需要说明的是，经过将所述碳排放数据输入至动力电池回收预测模型后，每一个碳排放数据的序列对应有一个动力电池回收量进行输出。
作为本实施例的一种可选方案，请参阅图2，在所述将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量之后，还包括步骤S103-S104：
S103：记录所述碳排放数据以及所述动力电池回收量，将所述碳排放数据以及所述动力电池回收量作为第二训练数据。
S104：根据所述第二训练数据，对所述动力电池回收预测模型进行再训练，输出再训练后的动力电池回收预测模型。
可以理解的是，通过记录所获取的碳排放数据以及预测得到的动力电池回收量，进而将新预测得到的数据作为第二训练数据，来对动力电池回收预测模型进行再训练，以使动力电池回收预测模型能够实现自动更新与修正，确保动力电池回收预测模型的准确性和精度，以使动力电池回收量预测精确度得到提高。
请参阅图3，所述动力电池回收预测模型的构建步骤，包括以下步骤S201-S202：
S201：获取动力电池的历史碳排放数据以及对应所述历史碳排放数据的动力电池回收数量，将所述历史碳排放数据以及所述历史碳排放数据对应的动力电池回收量作为第一训练数据。
所述动力电池的历史碳排放数据包括动力电池在生产阶段、运输阶段和使用阶段的历史碳排放数据。
需要说明的是，在本实施例中，动力电池在生产阶段、运输阶段和使用阶段的历史碳排放数据可以通过序列的形式进行获取，类似于步骤S101中的，动力电池碳排放数据的序列A＝[X1,X2,X3]，相应地，动力电池的历史碳排放数据的序列B＝[Y1,Y2,Y3]，其中，Y1为动力电池生产阶段所产生的历史碳排放数据，Y2为动力电池运输阶段所产生的历史碳排放数据，Y3为动力电池排放阶段所产生的历史碳排放数据。
对应历史碳排放数据的动力电池回收量，即为基于该碳排放数据下，进行 预测所得到的动力电池回收量的历史数据，即一个动力电池的历史碳排放数据的序列B＝[Y1,Y2,Y3]，对应于一个动力电池的历史回收量。
作为本实施例示例性的方案，在第一训练数据中选取70％的数据作为训练数据，15％的数据作为验证数据，15％的数据作为测试数据。
S202：根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型。
需要说明的是，在本实施例中，示例性地，请参阅图4，构建的初始动力电池回收预测模型为误差反向传播(Back Propagation，BP)神经网络模型，包括：输入层、隐藏层和输出层；其中，输入层中包括对应生产阶段、运输阶段和使用阶段的三个节点，输出层中包括对应动力电池回收数量的一个节点，隐藏层节点的数量根据实际的需求情况进行设定。
作为本实施例的一种可选方案，所述根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型，包括：构建初始动力电池回收预测模型；根据最优化算法，对所述初始动力电池回收预测模型的权值和阈值进行初始化操作；根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型。
需要说明的是，示例性地，本申请实施例中的最优化算法为莱文贝格－马夸特(levenberg-marquardt)算法，通常用于非线性最小二乘法的目标函数极小化。
在本实施例中，初始动力电池回收预测模型的权值和阈值μ进行初始化操作，示例性地，令k＝0，阈值μ＝μ0，其中常数μ0＝0.05。
可以理解的是，通过对初始动力电池回收预测模型进行构建，从而根据最优化算法来对初始动力电池回收预测模型的权值和阈值进行初始化操作，确保训练前的网络模型能够符合后续迭代训练的模型精度要求，并通过第一训练数据来对初始动力电池回收预测模型进行迭代训练，保证了训练得到的动力电池回收预测模型具备高精密度与高准确性，并且使得训练得到的动力电池回收预测模型避免了人为因素的干扰，使模型更加合理与客观。
作为本实施例的一种可选方案，所述根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型，包括：
对所述初始动力电池回收预测模型进行迭代训练，以使在每一次训练过程中，将根据输入至所述初始动力电池回收预测模型的所述第一训练数据中的历 史碳排放数据，计算得到输出值与所述第一训练数据中的动力电池回收量的误差指标函数，并根据输出值和误差指标函数，来计算并更新初始动力电池回收预测模型的权值和阈值，直至误差指标函数小于预设允许值时，输出训练完成后的动力电池回收预测模型。
在本实施例中，示例性地，计算模型输出及误差指标函数E(w)，其中，Yi为期望的模型输出向量；Y′i为实际的模型输出向量；P为样本数量；w为模型权值和阈值所组成的向量；ei(w)为误差。
进而计算误差矩阵J(w)，并计算Δw，Δw＝[JT(w)J(w)+μI]-1JT(w)e(w)；其中，I为单位矩阵；μ为用户定义的学习率，可选地，在本实施例中，μ设为0.05。
在本实施例中，可选地，预设允许值ε设为0.95，其中，预设允许值ε根据实际的需求与情况进行设定。若误差指标函数E(wk)＜ε＝0.95，则直接结束模型训练，并输出训练后的动力电池回收预测模型；否则，对模型的权值和阈值进行更新，以wk+1＝wk+Δw为新的权值和阈值向量，计算误差指标函数E(wk+1)，若E(wk+1)＜E(wk)，则令k＝k+1，阈值β(0＜β＜1)为0至1之间的随机数，其中k为迭代次数。
作为本实施例的一种可选方案，在输出训练完成后的动力电池回收预测模型之后，还需要通过输入测试集数据和验证集数据对动力电池回收预测模型的拟合残差和均方误差进行计算，以确保动力电池回收预测模型满足误差要求、拟合程度高以及线性化程度高。需要说明的是，均方误差是输出和目标输出之间的平均平方差，值越低越好，即零意味着没有误差。
可以理解的是，通过对初始动力电池回收预测模型进行迭代训练，确保了动力电池回收预测模型能够具备高精度和高准确性的预测能力，同时使得通过网络模型进行预测避免了现有人工预测的高复杂性，并在每一次训练过程中通过误差指标函数，来对网络模型的权值和阈值进行更新，使得初始动力电池回收预测模型逐渐收敛，以在误差指标函数小于预设允许值时，确保了能够准确且稳定地输出训练完成后的动力电池回收预测模型。
本申请实施例，通过对动力电池的历史碳排放数据进行采集，以及对应历史碳排放数据的动力电池回收数量作为训练数据，来对建立的初始动力电池回 收预测模型进行训练，从而能够准确得到动力电池回收预测模型，以使的后续可直接将动力电池的碳排放数据直接进行输入，即可准确输出动力电池的回收量，提高了企业工作人员对动力电池回收处理计划的制定效率，避免了现有人工主观进行预测的高复杂性，对电池回收量进行辅助决策，提高了动力电池回收的准确性与合理性。
实施例二
请参阅图5，本申请还提供一种动力电池回收量预测装置，包括：初始数据模块301、预测模块302和建模模块303。
所述初始数据模块301，设置为获取动力电池的碳排放数据。
所述预测模块302，设置为将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量。
所述建模模块303，包括：第一训练数据子模块3031和训练子模块3032。
所述第一训练数据子模块3031，设置为获取动力电池的历史碳排放数据以及对应所述历史碳排放数据的动力电池回收量，将所述历史碳排放数据以及所述历史碳排放数据对应的动力电池回收量作为第一训练数据。
所述训练子模块3032，设置为根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型。
作为本实施例的一种可选方案，所述训练子模块3032设置为通过以下方式根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型，包括：构建初始动力电池回收预测模型；根据最优化算法，对所述初始动力电池回收预测模型的权值和阈值进行初始化操作；根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型。
作为本实施例的一种可选方案，所述根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，从而输出训练完成后的动力电池回收预测模型，包括：对所述初始动力电池回收预测模型进行迭代训练，以使在每一次训练过程中，将根据输入至所述初始动力电池回收预测模型的所述第一训练数据中的历史碳排放数据，计算得到输出值与所述第一训练数据中的动力电池回收量的误差指标函数，并根据输出值和误差指标函数，来计算并更新初始动力电池回收预测模型网络的权值和阈值，直至误差指标函数小于预设允许值时，输出训练完成后的动力电池回收预测模型。
作为本实施例的一种可选方案，本实施例还包括：再训练模块304。
所述再训练模块304，设置为记录所述碳排放数据以及所述动力电池回收量，将所述碳排放数据以及所述动力电池回收量作为第二训练数据；根据所述第二训练数据，对所述动力电池回收预测模型进行再训练，输出再训练后的动力电池回收预测模型。
作为本实施例的一种可选方案，所述动力电池的碳排放数据包括动力电池在生产阶段、运输阶段和使用阶段的碳排放数据；所述动力电池的历史碳排放数据包括动力电池在生产阶段、运输阶段和使用阶段的历史碳排放数据。
所属领域的技术人员可以清楚的了解到，为描述的方便和简洁，上述描述的装置的工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
本申请实施例通过对动力电池的历史碳排放数据进行采集，以及对应历史碳排放数据的动力电池回收量作为训练数据，来对建立的初始动力电池回收预测模型进行训练，从而能够准确得到动力电池回收预测模型，以使的后续可直接将动力电池的碳排放数据直接进行输入，即可准确输出动力电池的回收量，提高了企业工作人员对动力电池回收处理计划的制定效率，避免了现有人工主观进行预测的高复杂性，对电池回收数量进行辅助决策，提高了动力电池回收的准确性与合理性。
实施例三
本申请还提供一种终端设备，包括：处理器、存储器以及存储在所述存储器中且被配置为由所述处理器执行的计算机程序，所述处理器执行所述计算机程序时实现如上任意一项实施例所述的动力电池回收量预测方法。
该实施例的终端设备包括：处理器、存储器以及存储在所述存储器中并可在所述处理器上运行的计算机程序、计算机指令。所述处理器执行所述计算机程序时实现上述实施例一中的多个步骤，例如图1所示的步骤S101至S102。或者，所述处理器执行所述计算机程序时实现上述装置实施例中多个模块/单元的功能，例如预测模块302。
示例性的，所述计算机程序可以被分割成一个或多个模块/单元，所述一个或者多个模块/单元被存储在所述存储器中，并由所述处理器执行，以完成本申请。所述一个或多个模块/单元可以是能够完成特定功能的一系列计算机程序指令段，该指令段用于描述所述计算机程序在所述终端设备中的执行过程。例如，所述预测模块302，设置为将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量。
所述终端设备可以是桌上型计算机、笔记本、掌上电脑及云端服务器等计算设备。所述终端设备可包括，但不仅限于，处理器、存储器。本领域技术人 员可以理解，示意图仅仅是终端设备的示例，并不构成对终端设备的限定，可以包括比图示更多或更少的部件，或者组合一些部件，或者不同的部件，例如所述终端设备还可以包括输入输出设备、网络接入设备、总线等。
所称处理器可以是中央处理单元(Central Processing Unit，CPU)，还可以是其他通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现成可编程门阵列(Field-Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等，所述处理器是所述终端设备的控制中心，利用多种接口和线路连接整个终端设备的多个部分。
所述存储器可设置为存储所述计算机程序和/或模块，所述处理器通过运行或执行存储在所述存储器内的计算机程序和/或模块，以及调用存储在存储器内的数据，实现终端设备的多种功能。所述存储器可主要包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需的应用程序等；存储数据区可存储根据移动终端的使用所创建的数据等。此外，存储器可以包括高速随机存取存储器，还可以包括非易失性存储器，例如硬盘、内存、插接式硬盘，智能存储卡(Smart Media Card,SMC)，安全数字(Secure Digital,SD)卡，闪存卡(Flash Card)、至少一个磁盘存储器件、闪存器件、或其他易失性固态存储器件。
所述终端设备集成的模块/单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本申请实现上述实施例方法中的全部或部分流程，也可以通过计算机程序来指令相关的硬件来完成，所述的计算机程序可存储于一计算机可读存储介质中，该计算机程序在被处理器执行时，可实现上述多个方法实施例的步骤。所述计算机程序包括计算机程序代码，所述计算机程序代码可以为源代码形式、对象代码形式、可执行文件或一些中间形式等。所述计算机可读介质可以包括：能够携带所述计算机程序代码的任何实体或装置、记录介质、U盘、移动硬盘、磁碟、光盘、计算机存储器、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、电载波信号、电信信号以及软件分发介质等。需要说明的是，所述计算机可读介质包含的内容可以根据司法管辖区内立法和专利实践的要求进行适当的增减，例如在一些司法管辖区，根据立法和专利实践，计算机可读介质不包括电载波信号和电信信号。
实施例四
相应地，本申请还提供一种计算机可读存储介质，所述计算机可读存储介 质包括存储的计算机程序，其中，在所述计算机程序运行时控制所述计算机可读存储介质所在设备执行如上任意一项实施例所述的动力电池回收量预测方法。

Claims (10)

  1. 一种动力电池回收量预测方法，包括：
    获取动力电池的碳排放数据；
    将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量；
    其中，构建所述动力电池回收预测模型，包括：
    获取所述动力电池的历史碳排放数据以及对应所述历史碳排放数据的动力电池回收量，将所述历史碳排放数据以及所述历史碳排放数据对应的动力电池回收量作为第一训练数据；
    根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型。
  2. 根据权利要求1所述的方法，其中，所述根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，从而输出得到训练后的动力电池回收预测模型，包括：
    构建初始动力电池回收预测模型；
    根据最优化算法，对所述初始动力电池回收预测模型的权值和阈值进行初始化操作；
    根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型。
  3. 根据权利要求2所述的方法，其中，所述根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型，包括：
    对所述初始动力电池回收预测模型进行迭代训练，以使在每一次训练过程中，将根据输入至所述初始动力电池回收预测模型的所述第一训练数据中的历史碳排放数据，计算得到输出值与所述第一训练数据中的动力电池回收量的误差指标函数，并根据所述输出值和所述误差指标函数，来计算并更新所述初始动力电池回收预测模型的权值和阈值，直至所述误差指标函数小于预设允许值时，输出训练完成后的动力电池回收预测模型。
  4. 根据权利要求1所述的方法，在所述将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量之后，还包括：
    记录所述碳排放数据以及所述动力电池回收量，将所述碳排放数据以及所述动力电池回收量作为第二训练数据；
    根据所述第二训练数据，对所述动力电池回收预测模型进行再训练，输出再训练后的动力电池回收预测模型。
  5. 根据权利要求1-4任意一项所述的方法，其中，所述动力电池的碳排放数据包括所述动力电池在生产阶段、运输阶段和使用阶段的碳排放数据；
    所述动力电池的历史碳排放数据包括所述动力电池在生产阶段、运输阶段和使用阶段的历史碳排放数据。
  6. 一种动力电池回收量预测装置，包括：初始数据模块、预测模块和建模模块；
    所述初始数据模块，设置为获取动力电池的碳排放数据；
    所述预测模块，设置为将所述碳排放数据输入至动力电池回收预测模型，输出得到动力电池回收量；
    所述建模模块，包括：第一训练数据子模块和训练子模块；
    所述第一训练数据子模块，设置为获取所述动力电池的历史碳排放数据以及对应所述历史碳排放数据的动力电池回收量，将将所述历史碳排放数据以及所述历史碳排放数据对应的动力电池回收量作为第一训练数据；
    所述训练子模块，设置为根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型。
  7. 根据权利要求6所述的装置，其中，所述训练子模块设置为通过以下方式根据所述第一训练数据，对构建的初始动力电池回收预测模型进行训练，输出训练后的动力电池回收预测模型：
    构建初始动力电池回收预测模型；
    根据最优化算法，对所述初始动力电池回收预测模型的权值和阈值进行初始化操作；
    根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型。
  8. 根据权利要求7所述的装置，其中，所述根据所述第一训练数据，对所述初始动力电池回收预测模型进行迭代训练，输出训练完成后的动力电池回收预测模型，包括：
    对所述初始动力电池回收预测模型进行迭代训练，以使在每一次训练过程中，将根据输入至所述初始动力电池回收预测模型的所述第一训练数据中的历史碳排放数据，计算得到输出值与所述第一训练数据中的动力电池回收量的误差指标函数，并根据所述输出值和所述误差指标函数，来计算并更新所述初始 动力电池回收预测模型的权值和阈值，直至所述误差指标函数小于预设允许值时，输出训练完成后的动力电池回收预测模型。
  9. 根据权利要求6所述的装置，还包括：再训练模块；
    所述再训练模块，设置为记录所述碳排放数据以及所述动力电池回收量，将所述碳排放数据以及所述动力电池回收量作为第二训练数据；根据所述第二训练数据，对所述动力电池回收预测模型进行再训练，输出再训练后的动力电池回收预测模型。
  10. 一种计算机可读存储介质，包括存储的计算机程序；其中，所述计算机程序在运行时控制所述计算机可读存储介质所在的设备执行如权利要求1-5中任一项所述的动力电池回收量预测方法。
PCT/CN2023/081916 2022-09-16 2023-03-16 动力电池回收量预测方法、装置及介质 WO2024055546A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211132754.3A CN115618998A (zh) 2022-09-16 2022-09-16 一种基于碳排放的动力电池回收量预测方法及装置
CN202211132754.3 2022-09-16

Publications (1)

Publication Number Publication Date
WO2024055546A1 true WO2024055546A1 (zh) 2024-03-21

Family

ID=84858201

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/081916 WO2024055546A1 (zh) 2022-09-16 2023-03-16 动力电池回收量预测方法、装置及介质

Country Status (2)

Country Link
CN (1) CN115618998A (zh)
WO (1) WO2024055546A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115618998A (zh) * 2022-09-16 2023-01-17 广东邦普循环科技有限公司 一种基于碳排放的动力电池回收量预测方法及装置
CN116664004B (zh) * 2023-06-20 2024-05-03 北京工业大学 一种三元动力电池回收碳减排效益分配方法及装置

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021184825A1 (zh) * 2020-03-20 2021-09-23 支付宝(杭州)信息技术有限公司 数据处理方法、装置、电子设备及存储介质
CN114004375A (zh) * 2021-10-27 2022-02-01 广东邦普循环科技有限公司 一种动力电池回收利用碳排放核算边界的界定方法和装置
CN114090939A (zh) * 2021-11-05 2022-02-25 国网湖北省电力有限公司经济技术研究院 一种储能电池全生命周期碳排放系数确定方法
CN114897226A (zh) * 2022-04-25 2022-08-12 广东邦普循环科技有限公司 一种碳排放量预测方法、装置、设备及存储介质
CN115049095A (zh) * 2022-03-23 2022-09-13 上海创能国瑞新能源科技股份有限公司 一种提升设备能效和碳排放动态优化系统及相关装置
CN115618998A (zh) * 2022-09-16 2023-01-17 广东邦普循环科技有限公司 一种基于碳排放的动力电池回收量预测方法及装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021184825A1 (zh) * 2020-03-20 2021-09-23 支付宝(杭州)信息技术有限公司 数据处理方法、装置、电子设备及存储介质
CN114004375A (zh) * 2021-10-27 2022-02-01 广东邦普循环科技有限公司 一种动力电池回收利用碳排放核算边界的界定方法和装置
CN114090939A (zh) * 2021-11-05 2022-02-25 国网湖北省电力有限公司经济技术研究院 一种储能电池全生命周期碳排放系数确定方法
CN115049095A (zh) * 2022-03-23 2022-09-13 上海创能国瑞新能源科技股份有限公司 一种提升设备能效和碳排放动态优化系统及相关装置
CN114897226A (zh) * 2022-04-25 2022-08-12 广东邦普循环科技有限公司 一种碳排放量预测方法、装置、设备及存储介质
CN115618998A (zh) * 2022-09-16 2023-01-17 广东邦普循环科技有限公司 一种基于碳排放的动力电池回收量预测方法及装置

Also Published As

Publication number Publication date
CN115618998A (zh) 2023-01-17

Similar Documents

Publication Publication Date Title
WO2024055546A1 (zh) 动力电池回收量预测方法、装置及介质
Wang et al. Modeling and optimization of multi-objective partial disassembly line balancing problem considering hazard and profit
Wang et al. A critical review of online battery remaining useful lifetime prediction methods
Mellit ANN-based GA for generating the sizing curve of stand-alone photovoltaic systems
CN105183994A (zh) 一种基于改进的i-elm的动力电池soc预测方法及装置
CN112491094B (zh) 一种混合驱动的微电网能量管理方法、系统及装置
Sun et al. A cloud-edge collaborative strategy for capacity prognostic of lithium-ion batteries based on dynamic weight allocation and machine learning
CN109713762A (zh) 一种充电控制方法、装置及终端设备
US20220393467A1 (en) Method and system for energy scheduling of shared energy storage considering degradation cost of energy storage
Zhang et al. Accurate prediction approach of SOH for lithium-ion batteries based on LSTM method
Li et al. A Comprehensive Review of Second Life Batteries Toward Sustainable Mechanisms: Potential, Challenges, and Future Prospects
CN114202229B (zh) 基于深度强化学习的微电网的能量管理策略的确定方法
CN112348352B (zh) 基于大数据分析的电力预算建议方案自动生成方法
Cong et al. Retired lithium-ion battery pack disassembly line balancing based on precedence graph using a hybrid genetic-firework algorithm for remanufacturing
WO2024060539A1 (zh) 动力电池的年度碳排放量估算方法及装置
CN115360768A (zh) 基于muzero和深度强化学习的电力调度方法、装置及存储介质
CN114142460B (zh) 综合能源系统中储能双层目标优化配置方法及终端
CN115864611A (zh) 储能电池安全储能管理方法、系统、设备及存储介质
CN116050250A (zh) 基于混合驱动模型的电网数字孪生模型建模方法及装置
CN114880929A (zh) 一种基于深度强化学习的多能流优化智能仿真方法及系统
Zhang et al. Novel Improved Particle Swarm Optimization-Extreme Learning Machine Algorithm for State of Charge Estimation of Lithium-Ion Batteries
CN112766609A (zh) 一种基于云计算的用电量预测方法
CN113644682A (zh) 高渗透有源配电网多台区协同管控方法、装置和终端设备
CN113298329A (zh) 训练、策略生成方法、系统、计算机装置及存储介质
US11709469B2 (en) Method for optimising the physical model of an energy installation and control method using such a model

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23864289

Country of ref document: EP

Kind code of ref document: A1