Energy consumption prediction method and device for discrete manufacturing workshop

Abstract

The invention relates to an energy consumption prediction method and device for a discrete manufacturing workshop, and belongs to the technical field of energy consumption prediction.The energy consumption prediction method for the discrete manufacturing workshop comprises the steps that process parameters, manual parameters, environment parameters and production target numbers of a target discrete manufacturing workshop are obtained to serve as an initial parameter set, the initial parameter set is preprocessed; taking the preprocessed initial parameter set as the input of an energy consumption prediction model, and determining the output of the energy consumption prediction model as the energy consumption prediction result of the target discrete manufacturing workshop; wherein the energy consumption prediction model is obtained by training on the basis of historical process parameters, historical manual parameters, historical environmental parameters, historical production target numbers and historical energy consumption data of the target discrete manufacturing workshop. According to the method, energy consumption prediction of the discrete manufacturing workshop is realized, and meanwhile, influence factors of energy consumption are fully considered, so that the accuracy of energy consumption prediction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy consumption prediction, and in particular to a method and device for predicting energy consumption in a discrete manufacturing workshop. Background Art

[0002] As the consumption and price of industrial energy continue to grow, many countries are increasingly concerned about sustainable economic development. As the material basis of the national economy, the manufacturing industry must maintain a high growth rate. In addition, the manufacturing industry is also a major area of ​​energy consumption. Therefore, energy management and control in the manufacturing industry is of great significance to achieving the goal of energy conservation and emission reduction.

[0003] The discrete manufacturing workshop is a large and complex system with many factors affecting its energy consumption. Due to the influence of these factors, the prediction of workshop energy consumption data is very complicated. Therefore, it is crucial to construct the manufacturing system in a structured, hierarchical and modular manner. However, methods such as artificial neural networks, decision trees, and support vector machines still require a lot of manual operations to extract features from the data, which increases the complexity of prediction. With the increase in data volume and the development of computers, workshop energy consumption prediction methods have gradually developed into deep learning, which can directly extract data features through the connection between neural network layers and can cope with large data sets. However, traditional deep learning algorithms are not sufficiently relevant to the processing of time series data, and the influencing factors considered in the prediction are not comprehensive enough, resulting in the lack of prediction accuracy in the existing discrete manufacturing workshop energy consumption prediction scheme. Summary of the invention

[0004] In view of this, it is necessary to provide a method and device for predicting energy consumption of a discrete manufacturing workshop, so as to solve the problem that the existing energy consumption prediction scheme for discrete manufacturing workshops is lacking in prediction accuracy.

[0005] In order to solve the above problems, the present invention provides an energy consumption prediction method for a discrete manufacturing workshop, comprising: Acquire process parameters, manual parameters, environmental parameters and production target numbers of a target discrete manufacturing workshop as an initial parameter set, and preprocess the initial parameter set; Using the preprocessed initial parameter set as input of an energy consumption prediction model, and determining the output of the energy consumption prediction model as an energy consumption prediction result of the target discrete manufacturing workshop; The energy consumption prediction model is trained based on historical process parameters, historical manual parameters, historical environmental parameters, historical production target numbers and historical energy consumption data of the target discrete manufacturing workshop.

[0006] In one possible implementation, the process parameters include the feed speed, spindle speed, cutting depth and processing material type of the machine tool, the manual parameters include the age, length of service and gender of the workers, and the environmental parameters include the temperature, humidity, wind speed and weather conditions of the target discrete manufacturing workshop.

[0007] In a possible implementation manner, preprocessing the initial parameter set includes: Supplementing missing values ​​in the initial parameter set to obtain a first parameter set; One-hot encodes the categorical variables in the first parameter set to obtain a second parameter set; Normalizing the second parameter set to obtain a third parameter set; Perform data dimension reduction processing on the third parameter set to obtain the preprocessed initial parameter set.

[0008] In a possible implementation, supplementing the initial parameter set with missing values ​​to obtain the first parameter set includes: The initial parameter set is supplemented with missing values ​​based on the average value of each quantitative variable in the initial parameter set to obtain the first parameter set.

[0009] In a possible implementation, performing one-hot encoding on the categorical variables in the first parameter set to obtain the second parameter set includes: The processing material type, the gender of the worker, and the weather condition in the first parameter set are one-hot encoded to obtain the second parameter set.

[0010] In one possible implementation, the energy consumption prediction model is trained with the historical process parameters, historical manual parameters, historical environmental parameters and historical production target numbers of the target discrete manufacturing workshop as input, the historical energy consumption data of the target discrete manufacturing workshop as labels, and the mean square error between the output of the energy consumption prediction model and the label as the loss function.

[0011] In a possible implementation, the energy consumption prediction model is a time recurrent neural network including 1 input layer, 1 hidden layer, 3 fully connected layers and 1 output layer.

[0012] The present invention also provides an energy consumption prediction device for a discrete manufacturing workshop, comprising: An acquisition module, used for acquiring process parameters, manual parameters, environmental parameters and production target numbers of a target discrete manufacturing workshop as an initial parameter set, and preprocessing the initial parameter set; A prediction module, used to use the preprocessed initial parameter set as an input of an energy consumption prediction model, and determine the output of the energy consumption prediction model as an energy consumption prediction result of the target discrete manufacturing workshop; The energy consumption prediction model is trained based on historical process parameters, historical manual parameters, historical environmental parameters, historical production target numbers and historical energy consumption data of the target discrete manufacturing workshop.

[0013] The present invention also provides an electronic device, including a memory and a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the energy consumption prediction method for discrete manufacturing workshops as described above is implemented.

[0014] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for predicting energy consumption of a discrete manufacturing workshop as described above is implemented.

[0015] The beneficial effects of the present invention are as follows: the energy consumption prediction method and device for a discrete manufacturing workshop provided by the present invention first obtain the process parameters, manual parameters, environmental parameters and production target numbers of the target discrete manufacturing workshop as an initial parameter set, thereby fully considering various factors affecting energy consumption, and then preprocessing the initial parameter set to reduce the calculation complexity of subsequent energy consumption prediction, and finally using the preprocessed initial parameter set as the input of an energy consumption prediction model, and predicting the energy consumption of the target discrete manufacturing workshop through the energy consumption prediction model. While realizing the energy consumption prediction of discrete manufacturing workshops, the present invention fully considers the factors affecting energy consumption, thereby improving the accuracy of energy consumption prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a flow chart of an embodiment of a method for predicting energy consumption of a discrete manufacturing workshop provided by the present invention; Figure 2 A schematic diagram of an embodiment of an energy consumption system of a discrete manufacturing workshop provided by the present invention; Figure 3 A schematic diagram of a process flow of an embodiment of an energy consumption prediction process for a discrete manufacturing workshop provided by the present invention; Figure 4 A schematic diagram of the structure of an embodiment of an energy consumption prediction device for a discrete manufacturing workshop provided by the present invention; Figure 5 A schematic structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0018] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0019] In the description of the present invention, reference to an "embodiment" means that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive of other embodiments. It is explicitly and implicitly understood by those skilled in the art that the described embodiments may be combined with other embodiments.

[0020] As the consumption and price of industrial energy continue to grow, many countries are increasingly concerned about sustainable economic development. As the material basis of the national economy, the manufacturing industry must maintain a high growth rate. In addition, the manufacturing industry is also a major area of ​​energy consumption. Therefore, energy management and control in the manufacturing industry is of great significance to achieving the goal of energy conservation and emission reduction.

[0021] The discrete manufacturing workshop is a large and complex system with many factors affecting its energy consumption. Due to the influence of these factors, the prediction of workshop energy consumption data is very complicated. Therefore, it is crucial to construct the manufacturing system in a structured, hierarchical and modular manner. However, methods such as artificial neural networks, decision trees, and support vector machines still require a lot of manual operations to extract features from the data, which increases the complexity of prediction. With the increase in data volume and the development of computers, workshop energy consumption prediction methods have gradually developed into deep learning, which can directly extract data features through the connection between neural network layers and can cope with large data sets. However, traditional deep learning algorithms are not sufficiently relevant to the processing of time series data, and the influencing factors considered in the prediction are not comprehensive enough, resulting in the lack of prediction accuracy in the existing discrete manufacturing workshop energy consumption prediction scheme.

[0022] In order to solve the above problems, the present invention provides an energy consumption prediction method for a discrete manufacturing workshop.

[0023] The specific embodiments are described in detail below: A specific embodiment of the present invention discloses a method for predicting energy consumption of a discrete manufacturing workshop, combining Figure 1 Come and see, Figure 1The schematic diagram of the process of an embodiment of the energy consumption prediction method for a discrete manufacturing workshop provided by the present invention includes step S101 and step S102, wherein: In step S101, the process parameters, manual parameters, environmental parameters and production target number of the target discrete manufacturing workshop are obtained as an initial parameter set, and the initial parameter set is preprocessed; In step S102, the preprocessed initial parameter set is used as the input of the energy consumption prediction model, and the output of the energy consumption prediction model is determined as the energy consumption prediction result of the target discrete manufacturing workshop; The energy consumption prediction model is trained based on historical process parameters, historical manual parameters, historical environmental parameters, historical production target numbers and historical energy consumption data of the target discrete manufacturing workshop.

[0024] During implementation, we can first obtain the process parameters, manual parameters, environmental parameters and production target number of the target discrete manufacturing workshop as the initial parameter set, and comprehensively consider the impact of various factors on energy consumption to improve the accuracy of energy consumption prediction.

[0025] After the initial parameter set is obtained, the initial parameter set may be preprocessed so as to be used for subsequent energy consumption prediction and further improve the accuracy of energy consumption prediction.

[0026] Then, the preprocessed initial parameter set can be used as the input of the energy consumption prediction model, and then the output of the energy consumption prediction model can be determined as the energy consumption prediction result of the target discrete manufacturing workshop to achieve the energy consumption prediction of the target discrete manufacturing workshop.

[0027] The energy consumption prediction model can be trained based on the historical process parameters, historical manual parameters, historical environmental parameters, historical production target numbers and historical energy consumption data of the target discrete manufacturing workshop.

[0028] The energy consumption prediction method for a discrete manufacturing workshop provided by the present invention can be applied to energy consumption prediction scenarios of discrete manufacturing workshops, and can also be applied to energy consumption prediction scenarios of other productions, and the present invention does not make specific limitations on this.

[0029] Compared with the prior art, the energy consumption prediction method for a discrete manufacturing workshop provided in this embodiment first obtains the process parameters, manual parameters, environmental parameters and production target numbers of the target discrete manufacturing workshop as the initial parameter set, thereby fully considering various factors affecting energy consumption, and then preprocesses the initial parameter set to reduce the calculation complexity of subsequent energy consumption prediction. Finally, the preprocessed initial parameter set is used as the input of the energy consumption prediction model, and the energy consumption of the target discrete manufacturing workshop is predicted through the energy consumption prediction model. While realizing the energy consumption prediction of discrete manufacturing workshops, the present invention fully considers the factors affecting energy consumption, thereby improving the accuracy of energy consumption prediction.

[0030] Exemplarily, the process parameters include the feed speed, spindle speed, cutting depth and processing material type of the machine tool, the manual parameters include the age, length of service and gender of the workers, and the environmental parameters include the temperature, humidity, wind speed and weather conditions of the target discrete manufacturing workshop.

[0031] Specifically, the process parameters of the target discrete manufacturing workshop may include the feed speed, spindle speed, cutting depth and processing material type of the machine tool; the manual parameters of the target discrete manufacturing workshop may include the age, length of service and gender of the workers; and the environmental parameters of the target discrete manufacturing workshop may include the temperature, humidity, wind speed and weather conditions of the workshop.

[0032] Exemplarily, the preprocessing the initial parameter set includes: Supplementing missing values ​​in the initial parameter set to obtain a first parameter set; One-hot encodes the categorical variables in the first parameter set to obtain a second parameter set; Normalizing the second parameter set to obtain a third parameter set; Perform data dimension reduction processing on the third parameter set to obtain the preprocessed initial parameter set.

[0033] Specifically, when preprocessing the initial parameter set, the missing values ​​of the initial parameter set may be supplemented first to obtain a first parameter set to ensure the quality of the data.

[0034] Next, in order to ensure that the parameter set can be used for energy consumption prediction, the categorical variables in the first parameter set can be one-hot encoded to obtain the second parameter set.

[0035] Then, in order to reduce the influence of different dimensions on the prediction accuracy, the second parameter set may be normalized, such as maximum and minimum normalization, to obtain a third parameter set.

[0036] Finally, in order to reduce the amount of calculation in energy consumption prediction, the third parameter set may be subjected to data dimensionality reduction processing, such as principal component analysis (PCA) dimensionality reduction, to obtain a preprocessed initial parameter set.

[0037] Exemplarily, supplementing the missing values ​​of the initial parameter set to obtain the first parameter set includes: The initial parameter set is supplemented with missing values ​​based on the average value of each quantitative variable in the initial parameter set to obtain the first parameter set.

[0038] Specifically, when supplementing missing values ​​in the initial parameter set, the average value of each quantitative variable in the initial parameter set can be calculated first, and then the average value of each quantitative variable can be supplemented to the corresponding missing value to complete the supplement of the missing values ​​of the initial parameter set and obtain the first parameter set.

[0039] Exemplarily, performing one-hot encoding on the categorical variables in the first parameter set to obtain the second parameter set includes: The processing material type, the gender of the worker, and the weather condition in the first parameter set are one-hot encoded to obtain the second parameter set.

[0040] Specifically, when one-hot encoding is performed on the categorical variables in the first parameter set, the difficult-to-quantify features in the first parameter set, such as the type of processed materials, the gender of the workers, and the weather conditions, can be one-hot encoded to obtain the second parameter set.

[0041] Exemplarily, the energy consumption prediction model is trained with historical process parameters, historical manual parameters, historical environmental parameters and historical production target numbers of the target discrete manufacturing workshop as input, historical energy consumption data of the target discrete manufacturing workshop as labels, and the mean square error between the output of the energy consumption prediction model and the label as the loss function.

[0042] Specifically, when training the energy consumption prediction model, the historical process parameters, historical manual parameters, historical environmental parameters, and historical production target numbers of the target discrete manufacturing workshop can be used as the input of the model, and the historical energy consumption data of the target discrete manufacturing workshop can be used as labels to train the model. At the same time, when training the model, the mean-square error (MSE) between the output of the energy consumption prediction model and the label can be used as a loss function to optimize the model parameters. In addition, MSE and mean absolute error (MAE) can also be used to evaluate the performance of the model.

[0043] Exemplarily, the energy consumption prediction model is a time recurrent neural network comprising 1 input layer, 1 hidden layer, 3 fully connected layers and 1 output layer.

[0044] Specifically, the energy consumption prediction model described in the above content can be a time recurrent neural network including 1 input layer, 1 hidden layer, 3 fully connected layers and 1 output layer, such as a long short-term memory (LSTM) network, or a bidirectional recurrent neural network (Bi-RNN).

[0045] The following is a specific application scenario to better illustrate the technical solution of the present invention: Combination Figure 2 Come and see, Figure 2 The present invention provides a schematic diagram of an embodiment of an energy consumption system for a discrete manufacturing workshop. The present invention proposes an improved method for predicting workshop energy consumption that takes environmental factors, process parameters, and workshop equipment into comprehensive consideration. This method can more accurately predict the energy consumption of a discrete manufacturing workshop, guide the scheduling of energy consumption of each part of the workshop through energy consumption prediction, and reduce the energy cost of the workshop. The present invention develops a comprehensive indicator system for discrete workshops from three levels: process parameters, equipment, and environment, and uses these characteristics as inputs to the model to improve the accuracy of the model. The present invention can also be used to build an energy consumption management system for discrete manufacturing workshops, realize real-time collection of workshop energy consumption information, and provide data support for the coordinated scheduling of multiple energy sources on the energy supply side of subsequent workshops.

[0046] The process parameter layer refers to the impact of different processing parameters of the workpiece on the energy consumption of the workshop. The process parameters mainly consider the three elements of cutting (cutting depth, feed speed and spindle speed) and the type of tool. Different process parameters affect the power of the machine tool, and then affect the energy consumption of the entire workshop. By analyzing and collecting the process parameters of each step, all the steps to complete the processing of the product are integrated and analyzed for their energy consumption.

[0047] The equipment layer refers to the equipment used in the process of processing raw materials into products according to production requirements, including CNC machine tools, conveyor belts, robotic arms, forklifts, air conditioners and other equipment. Machine tools are the main energy-consuming units in discrete manufacturing workshops. For the same production task, the different types and models of machine tools will directly affect the amount of energy consumption. Equipment such as conveyor belts, robotic arms and forklifts can be collectively referred to as auxiliary equipment. Their main purpose is to ensure that the machine tools can be processed normally according to the production arrangement and the appropriate production environment. These energy consumption data are basically fixed values ​​and can be directly obtained from the historical energy consumption of the workshop.

[0048] The environment of the workshop includes the worker's operating proficiency and gender, atmospheric temperature and humidity, whether it is a holiday, workshop scheduling, etc. In the production and processing of discrete workshops, the worker's proficiency can be expressed by the worker's job length of service. The workers who have been working in the same position for a long time have a large difference in operating proficiency compared with new workers, and their production efficiency is also very different. The gender ratio of workers has a greater impact on some positions. In some positions that require physical strength, men have more advantages than women. Therefore, considering the workers' length of service and gender in the energy consumption impact of discrete workshops can improve the accuracy of energy consumption assessment. In addition, the gender of the workers needs to be one-hot encoded, with males represented as 1 and females represented as 0. Factors such as temperature, humidity and wind speed can be summarized as external factors. Different weather factors can directly lead to the operation of the workshop lighting system, ventilation system and humidification system, thereby affecting the energy consumption of the entire workshop. In addition, for special processing tasks, it is necessary to ensure that certain environmental conditions can be guaranteed before processing, so weather factors are one of the factors affecting energy consumption. In the production process of discrete manufacturing workshops, due to the large number of processing equipment and complex product processing procedures, reasonable workshop production scheduling arrangements can reduce the idle time of equipment, thereby reducing the time for product completion and improving production efficiency. Therefore, reasonable workshop production scheduling arrangements are of great significance to energy conservation and emission reduction in discrete workshops.

[0049] Combination Figure 3 Come and see, Figure 3 This is a flow chart of an embodiment of the energy consumption prediction process of a discrete manufacturing workshop provided by the present invention, and the specific process is as follows: Collect energy consumption data of workshop equipment to form a data set. Since the data collection device may have abnormal conditions, resulting in missing and abnormal data sets, in order to ensure the quality of the data set, fill the missing values ​​of the data set and use the average value of its features to fill the missing values, thereby ensuring the quality of the data. Since the holiday feature is a nominal feature, the feature is one-hot encoded to ensure the normal use of subsequent data sets.

[0050] After one-hot encoding of the original data set, in order to reduce the impact of different dimensions on the accuracy of the model, the original data is normalized so that the data is at the same dimension level. The normalization formula is:

[0051] In the formula is the normalized data, is the original data, is the minimum value of the original data, is the maximum value of the original data.

[0052] There are multiple features in the present invention. Since there may be correlations between their own characteristics, the training time and time cost of model training may be increased, and there may even be a risk of overfitting or falling into a local minimum. Therefore, the data is processed by dimensionality reduction to reduce the amount of calculation of the model. First, the covariance of the matrix composed of the data set is calculated, and then the eigenvalue decomposition of the covariance matrix is ​​performed to select the K eigenvectors corresponding to the largest K eigenvalues. Finally, the original data is projected onto the selected eigenvectors, and the projected data set is output. Calculate the contribution rate and cumulative contribution rate of each principal component, express most of the characteristics of the original data through low-dimensional data, and obtain a new matrix by reducing the dimensionality of the original data.

[0053] The present invention mainly studies the energy consumption of discrete workshops. In order to more accurately predict the energy consumption of workshops, the quality of input features must be ensured. Machine tools are the main energy-consuming units in workshops. In order to study the energy consumption of machine tools, the present invention starts with process parameters, expresses the energy consumption of machine tools from the three cutting elements of each process, and finally superimposes the energy consumption of each machine tool to complete the process to obtain the total energy consumption of the workshop machine tools. Suppose the workshop has a batch of orders of quantity W, the number of workshop machine tools is L, and it takes m processes to complete the processing of a product. The speed of the machine tool in each process is , the feed speed is , cutting depth is , then the total energy consumption of the machine tools in the workshop is:

[0054] Indicates Energy consumption of each process.

[0055] The energy consumption of other equipment in the workshop, such as conveyors, ventilation equipment, lighting equipment, dust removal equipment, etc., can be obtained from the historical data of the workshop. The energy consumption of these equipment is mainly affected by environmental factors, such as temperature, humidity, wind speed, weather conditions, etc., so these influencing factors can be used as inputs to the model.

[0056] In order to more accurately predict the energy consumption of the workshop, the present invention uses the three cutting elements, temperature, humidity and wind speed of each process as the input of the model, and the energy consumption of the discrete workshop as the output of the model, so as to predict the energy consumption of the machine tool. Since there are many machine tools in the workshop, the dimension of the input data may be too large, resulting in a high running speed and difficulty of model training. Therefore, the PCA dimensionality reduction method is used to link the input data and express most of the information of the input data with low-dimensional data.

[0057] The temporal characteristics of the data are learned through LSTM, and then regression prediction is achieved through the fully connected layer. The activation functions of the LSTM layer and the fully connected layer use the commonly used Relu function. Data with input less than 0 is directly discarded, and data with input greater than 0 is retained. The derivative of the Relu function performs well, either keeping the input or disappearing, greatly alleviating the problem of gradient disappearance in the neural network. The output of the model is a fully connected layer with 1 neuron, without the need to use an activation function, and then outputs an accurate prediction value.

[0058] For the prediction results of the algorithm, MSE and MAE are used to evaluate the performance of the model, and MSE is used to represent the loss function of the model. The loss function and average error formula are:

[0059]

[0060] in, is the predicted energy consumption value, is the actual energy consumption value, is the sample size.

[0061] The number of time steps is an important parameter of the model, which determines the time series length of the model input when performing energy consumption prediction. The selection of this parameter should consider both the integrity of the input data and the duration of the model training. On the one hand, a time series length that is too small will lead to the loss of model input information, resulting in a reduction in the accuracy of the prediction model. On the other hand, a time series length that is too long increases the complexity of the model, increasing the speed and cost of training. Therefore, the appropriate number of time steps is conducive to improving the accuracy of the model and can reduce the complexity of training. The present invention uses a grid search algorithm to optimize the model parameters until the optimal parameter input is found. It is finally determined that the accuracy of the model reaches the optimal when the number of time windows is set to 15.

[0062] The present invention also evaluates the prediction effect of the model through experiments. Different types of machine tools, ventilation equipment, and mechanical arms form a processing workshop. Clamp-on transformers, temperature and humidity sensors, Kistler 9257B three-component dynamometers, and WT1800 power meters are used to collect and record the historical data of the discrete manufacturing workshop. The processing material of the CNC machine tool is T300 laminate, and the tool type is 8cm and 10cm carbide high-strength drilling drill bits. The specific parameters of the field recording experiment collection device and equipment are shown in Table 1: Table 1: Discrete workshop equipment parameters

[0063] In this experiment, there are 12 input features, namely, the process parameters of each process in the workshop (spindle speed, feed speed, cutting depth, material type), environmental factors (temperature, humidity, wind speed, cloudy and sunny), personnel factors (gender, age, length of service), and the target production quantity of the day. The output feature is the total energy consumption of all equipment in the workshop, as shown in Table 2 below. Since there may be correlations between their own characteristics, the training time and time cost of model training may increase, and there may even be risks of overfitting or falling into local minima, so the data is reduced in dimension. Most of the features of the original data are expressed through low-dimensional data. When the original data is reduced in dimension to obtain a new matrix, when the number of principal components reaches 10, the cumulative contribution rate exceeds 95%. Therefore, the original matrix can be reduced in dimension to generate a matrix containing 10 principal components and no correlation between each principal component, and the new matrix composed of this feature is used as the new input matrix.

[0064] Table 2: Model input and output parameters

[0065] According to the data set collected in the previous article, 80% of the data set is used as the training set, and 20% is used as the test set. The LSTM neural network parameters are set as follows: the model includes an input layer, 1 hidden layer, 3 fully connected layers, and an output layer. The number of neurons in the input layer is 64, the number of neurons in the hidden layer is 32, the number of neurons in the fully connected layer is 32, 16, and 8, and the number of neurons in the output layer is 1. Secondly, the number of time steps and the number of samples passed each time of the model are obtained by the grid optimization algorithm to be 15 and 16 respectively. The number of model iterations is set to 130. After the number of iterations reaches 120 times, the loss function of the model tends to a stable value. The calculated MAE and MSE errors without considering environmental factors are 3.72% and 9.135% respectively. The MAE and MSE of the test set on this model are 2.406% and 6.883% respectively. The experiment shows that considering environmental factors in the factors affecting energy consumption can improve the accuracy of workshop energy consumption prediction. Moreover, the predicted energy consumption value taking environmental factors into account is very close to the actual test value, which can basically meet the allowable error range in actual production.

[0066] The embodiment of the present invention also provides an energy consumption prediction device for a discrete manufacturing workshop, Figure 4 Come and see, Figure 4 This is a schematic diagram of the structure of an embodiment of an energy consumption prediction device for a discrete manufacturing workshop provided by the present invention. The energy consumption prediction device 400 for a discrete manufacturing workshop includes: An acquisition module 401 is used to acquire process parameters, manual parameters, environmental parameters and production target numbers of a target discrete manufacturing workshop as an initial parameter set, and preprocess the initial parameter set; A prediction module 402, configured to use the preprocessed initial parameter set as an input of an energy consumption prediction model, and determine the output of the energy consumption prediction model as an energy consumption prediction result of the target discrete manufacturing workshop; The energy consumption prediction model is trained based on historical process parameters, historical manual parameters, historical environmental parameters, historical production target numbers and historical energy consumption data of the target discrete manufacturing workshop.

[0067] The specific implementation methods of each module of the energy consumption prediction device for discrete manufacturing workshops can refer to the description of the energy consumption prediction method for discrete manufacturing workshops mentioned above, and have similar beneficial effects, which will not be repeated here.

[0068] The embodiment of the present invention further provides an electronic device, Figure 5 Come and see, Figure 5 This is a structural diagram of an embodiment of an electronic device provided by the present invention. The electronic device 500 includes a processor 501, a memory 502, and a computer program stored in the memory 502 and executable on the processor 501. When the processor 501 executes the program, the energy consumption prediction method for a discrete manufacturing workshop as described above is implemented.

[0069] As a preferred embodiment, the electronic device 500 further includes a display 503 for displaying the energy consumption prediction method for discrete manufacturing workshops as executed by the processor 501 .

[0070] Exemplarily, the computer program may be divided into one or more modules / units, one or more modules / units are stored in the memory 502, and are executed by the processor 501 to complete the present invention. One or more modules / units may be a series of computer program instruction segments that can complete specific functions, and the instruction segments are used to describe the execution process of the computer program in the electronic device 500. For example, the computer program may be divided into the acquisition module 401 and the prediction module 402 in the above embodiment, and the specific functions of each module are as described above, and are not described one by one here.

[0071] The electronic device 500 may be a desktop computer, a notebook, a PDA or a smart phone with an adjustable camera module.

[0072] Among them, the processor 501 may be an integrated circuit chip with signal processing capabilities. The above-mentioned processor 501 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0073] The memory 502 may be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc. The memory 502 is used to store programs, and the processor 501 executes the program after receiving an execution instruction. The process definition method disclosed in any of the above embodiments of the present invention may be applied to the processor 501, or implemented by the processor 501.

[0074] The display 503 may be an LCD display screen or an LED display screen, for example, a display screen on a mobile phone.

[0075] Understandably, Figure 5 The structure shown is only a schematic diagram of the structure of the electronic device 500. The electronic device 500 may also include Figure 5 More or fewer components as shown. Figure 5 Each component shown in the figure can be implemented by hardware, software or a combination thereof.

[0076] The electronic device provided according to the above-mentioned embodiment of the present invention can be implemented with reference to the specific description of the energy consumption prediction method for discrete manufacturing workshops as described above according to the present invention, and has similar beneficial effects as the energy consumption prediction method for discrete manufacturing workshops as described above, which will not be repeated here.

[0077] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the energy consumption prediction method for a discrete manufacturing workshop as described above is implemented.

[0078] Generally speaking, the computer instructions for implementing the method of the present invention can be carried by any combination of one or more computer-readable storage media. Non-transitory computer-readable storage media can include any computer-readable media except for the signal itself that is temporarily propagating.

[0079] Computer-readable storage media may be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more conductors, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium that contains or stores a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0080] Computer program codes for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages, in particular, Python suitable for neural network computing and platform frameworks based on TensorFlow, PyTorch, etc. may be used. The program code may be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0081] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0082] The present invention discloses a method and device for predicting energy consumption of a discrete manufacturing workshop. Firstly, the process parameters, manual parameters, environmental parameters and production target number of a target discrete manufacturing workshop are obtained as an initial parameter set, thereby fully considering various factors affecting energy consumption. Then, the initial parameter set is preprocessed to reduce the calculation complexity of subsequent energy consumption prediction. Finally, the preprocessed initial parameter set is used as the input of an energy consumption prediction model. The energy consumption of the target discrete manufacturing workshop is predicted through the energy consumption prediction model. The present invention fully considers the factors affecting energy consumption while realizing the energy consumption prediction of discrete manufacturing workshops, thereby improving the accuracy of energy consumption prediction.

[0083] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for predicting energy consumption of a discrete manufacturing workshop, characterized in that: include: Acquire process parameters, manual parameters, environmental parameters and production target numbers of a target discrete manufacturing workshop as an initial parameter set, and preprocess the initial parameter set; Using the preprocessed initial parameter set as input of an energy consumption prediction model, and determining the output of the energy consumption prediction model as an energy consumption prediction result of the target discrete manufacturing workshop; The energy consumption prediction model is trained based on historical process parameters, historical manual parameters, historical environmental parameters, historical production target numbers and historical energy consumption data of the target discrete manufacturing workshop.

2. The energy consumption prediction method for discrete manufacturing workshop according to claim 1 is characterized in that: The process parameters include the feed speed, spindle speed, cutting depth and processing material type of the machine tool, the manual parameters include the age, length of service and gender of the workers, and the environmental parameters include the temperature, humidity, wind speed and weather conditions of the target discrete manufacturing workshop.

3. The energy consumption prediction method for discrete manufacturing workshop according to claim 2 is characterized in that: The preprocessing of the initial parameter set includes: Supplementing missing values ​​in the initial parameter set to obtain a first parameter set; One-hot encodes the categorical variables in the first parameter set to obtain a second parameter set; Normalizing the second parameter set to obtain a third parameter set; Perform data dimension reduction processing on the third parameter set to obtain the preprocessed initial parameter set.

4. The energy consumption prediction method for discrete manufacturing workshops according to claim 3 is characterized in that: The step of supplementing the missing values ​​of the initial parameter set to obtain a first parameter set includes: The initial parameter set is supplemented with missing values ​​based on the average value of each quantitative variable in the initial parameter set to obtain the first parameter set.

5. The energy consumption prediction method for discrete manufacturing workshop according to claim 3 is characterized in that: The one-hot encoding of the categorical variables in the first parameter set to obtain a second parameter set includes: The processing material type, the gender of the worker, and the weather condition in the first parameter set are one-hot encoded to obtain the second parameter set.

6. The method for predicting energy consumption of a discrete manufacturing workshop according to any one of claims 1 to 5, characterized in that: The energy consumption prediction model is trained with the historical process parameters, historical manual parameters, historical environmental parameters and historical production target numbers of the target discrete manufacturing workshop as input, the historical energy consumption data of the target discrete manufacturing workshop as labels, and the mean square error between the output of the energy consumption prediction model and the label as the loss function.

7. The method for predicting energy consumption of a discrete manufacturing workshop according to any one of claims 1 to 5, characterized in that: The energy consumption prediction model is a time recurrent neural network including 1 input layer, 1 hidden layer, 3 fully connected layers and 1 output layer.

8. An energy consumption prediction device for a discrete manufacturing workshop, characterized in that: include: An acquisition module, used for acquiring process parameters, manual parameters, environmental parameters and production target numbers of a target discrete manufacturing workshop as an initial parameter set, and preprocessing the initial parameter set; A prediction module, used to use the preprocessed initial parameter set as an input of an energy consumption prediction model, and determine the output of the energy consumption prediction model as an energy consumption prediction result of the target discrete manufacturing workshop; The energy consumption prediction model is trained based on historical process parameters, historical manual parameters, historical environmental parameters, historical production target numbers and historical energy consumption data of the target discrete manufacturing workshop.

9. An electronic device, characterized in that: The invention comprises a memory and a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the energy consumption prediction method for a discrete manufacturing workshop according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the method for predicting energy consumption of a discrete manufacturing workshop as claimed in any one of claims 1 to 7 is implemented.

Images