Flue gas oxygen content load prediction method and device based on joint learning

Abstract

The present invention provides a method and device for predicting the load of flue gas oxygen content based on joint learning. The method comprises: determining the sample weights corresponding to the data sets of multiple local equipment groups and the target equipment data sets respectively by the participating parties; training the prediction neural network models of multiple equipment groups according to the data sets of multiple local equipment groups and the corresponding sample weights; uploading the prediction neural network models of multiple local equipment groups to the central node for model aggregation to obtain the aggregated prediction neural network model; training the aggregated prediction neural network model according to preset training conditions to obtain a joint prediction model; predicting the smoke oxygen content value of the target equipment according to the joint prediction model and the sample weights corresponding to the target equipment data set. The present invention solves the problem of resource waste caused by the inability to improve the accuracy of the prediction of the load of flue gas oxygen content of energy equipment in the prior art.

Description

Technical Field

[0001] The present disclosure relates to the field of integrated energy technology, and in particular to a method and device for predicting flue gas oxygen content load based on joint learning. Background Art

[0002] With the widespread application of integrated energy, thermal efficiency is an important indicator for measuring gas boilers. Generally, the oxygen content of boiler flue gas is controlled to the optimal design value to achieve the highest thermal efficiency under the operation of different energy equipment. The oxygen content of flue gas is usually measured and maintained by zirconium oxide measuring instruments, but the cost is relatively high.

[0003] For example, in the field of distributed energy, small gas boilers generally give up installing zirconia measuring instruments in order to save costs, which makes it impossible to achieve closed-loop control and optimal thermal efficiency operation. In particular, when the calorific value of the gas is unstable, the thermal efficiency will be mostly sacrificed. That is, the data distribution of different boilers is different, which greatly affects the prediction accuracy of the data, so that the prediction of the oxygen content of boiler flue gas load will not be accurate, which brings great economic losses to enterprises or factories that use energy equipment, so it is urgent to solve this problem. Summary of the invention

[0004] In view of this, the embodiments of the present disclosure provide a flue gas oxygen content load prediction method, device, computer equipment and computer-readable storage medium based on joint learning to solve the problem of resource waste caused by the inability to improve the accuracy of flue gas oxygen content load prediction of energy equipment in the prior art.

[0005] A first aspect of an embodiment of the present disclosure provides a flue gas oxygen content load prediction method based on joint learning, comprising:

[0006] The participants determine the sample weights corresponding to the data sets of the local multiple groups of devices and the data sets of the target devices respectively;

[0007] According to the data sets of multiple groups of local devices and the corresponding sample weights, a prediction neural network model for multiple groups of devices is trained;

[0008] Upload the prediction neural network models of multiple groups of local devices to the central node for model aggregation to obtain the aggregated prediction neural network model;

[0009] Training the aggregated prediction neural network model according to preset training conditions to obtain a joint prediction model;

[0010] According to the joint prediction model and the sample weights corresponding to the target equipment data set, the oxygen content value of the flue gas of the target equipment is predicted.

[0011] A second aspect of the embodiment of the present disclosure provides a flue gas oxygen content load prediction device based on joint learning, comprising:

[0012] A determination module, used by the participants to determine the sample weights corresponding to the data sets of multiple local devices and the target device data sets respectively;

[0013] A first training module is used to train a prediction neural network model for multiple groups of devices based on the data sets of multiple groups of local devices and corresponding sample weights;

[0014] An aggregation module is used to upload the prediction neural network models of multiple groups of local devices to the central node for model aggregation to obtain an aggregated prediction neural network model;

[0015] The second training module trains the aggregated prediction neural network model according to preset training conditions to obtain a joint prediction model;

[0016] The prediction module is used to predict the oxygen content value of the flue gas of the target device according to the joint prediction model and the sample weight corresponding to the target device data set.

[0017] According to a third aspect of an embodiment of the present disclosure, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0018] According to a fourth aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0019] Compared with the prior art, the disclosed embodiment has the following beneficial effects: the data sets of the local multiple groups of equipment and the sample weights corresponding to the target equipment data sets are determined by the participants respectively; the prediction neural network models of the multiple groups of equipment are trained according to the data sets of the local multiple groups of equipment and the corresponding sample weights; the prediction neural network models of the local multiple groups of equipment are uploaded to the central node for model aggregation to obtain the aggregated prediction neural network model; the aggregated prediction neural network model is trained according to the preset training conditions to obtain the joint prediction model; the oxygen content value of the flue gas of the target equipment is predicted according to the sample weights corresponding to the joint prediction model and the target equipment data set. The disclosed embodiment solves the problem of resource waste caused by the inability to improve the accuracy of the prediction of the oxygen content load of the flue gas of the energy equipment in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 is a schematic diagram of a joint learning architecture of an embodiment of the present disclosure;

[0022] Figure 2 is a flow chart of a flue gas oxygen content load prediction method based on joint learning provided by an embodiment of the present disclosure;

[0023] Figure 3 is a block diagram of a flue gas oxygen content load prediction device based on joint learning provided by an embodiment of the present disclosure;

[0024] Figure 4 It is a schematic diagram of a computer device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0025] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present disclosure. However, it should be clear to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obstructing the description of the present disclosure with unnecessary details.

[0026] Federated learning refers to the comprehensive use of various AI (Artificial Intelligence) technologies to jointly explore the value of data and give rise to new intelligent formats and models based on joint modeling, while ensuring data security and user privacy. Federated learning has at least the following characteristics:

[0027] (1) A weakly centralized joint training model in which participating nodes control their own data to ensure data privacy and security in the process of co-creating intelligence.

[0028] (2) In different application scenarios, we use screening and / or combination of AI algorithms and privacy-preserving computing to establish multiple model aggregation optimization strategies to obtain high-level, high-quality models.

[0029] (3) Under the premise of ensuring data security and user privacy, methods to improve the efficiency of federated learning are obtained based on multiple model aggregation optimization strategies. The efficiency method can be achieved by solving problems including computing architecture parallelism, information interaction in large-scale cross-domain networks, intelligent perception, and exception handling mechanisms to improve the overall efficiency of federated learning.

[0030] (4) Obtain the needs of multiple users in each scenario, determine a reasonable assessment of the true contribution of each joint participant through a mutual trust mechanism, and distribute incentives.

[0031] Based on the above methods, we can establish an AI technology ecosystem based on federated learning, give full play to the value of industry data, and promote the implementation of scenarios in vertical fields.

[0032] A flue gas emission prediction method and device based on joint learning according to an embodiment of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0033] Figure 1 Schematic diagram of a joint learning architecture of an embodiment of the present disclosure. Figure 1 As shown, the architecture of the federated learning may include a server (central node) 101 and participants 102, 104, and 104. A participant may be composed of one or more clients.

[0034] In the joint learning process, the basic model can be established by the server 101, and the server 101 sends the model to the participants 102, 104, and 104 with which the communication connection is established. The basic model can also be uploaded to the server 101 by any participant after establishment, and the server 101 sends the model to other participants with which the communication connection is established. The participants 102, 104, and 104 build a model according to the downloaded basic structure and model parameters, use local data to train the model, obtain updated model parameters, and encrypt and upload the updated model parameters to the server 101. The server 101 aggregates the model parameters sent by the participants 102, 104, and 104 to obtain global model parameters, and transmits the global model parameters back to the participants 102, 104, and 104. The participants 102, 104, and 104 iterate their respective models according to the received global model parameters until the model finally converges, thereby realizing the training of the model. During the joint learning process, the data uploaded by participants 102, 104 and 104 are model parameters, local data will not be uploaded to server 101, and all participants can share the final model parameters, so joint modeling can be achieved while ensuring data privacy.

[0035] It should be noted that the number of participants is not limited to three as described above, but can be set as needed, and the embodiments of the present disclosure do not limit this.

[0036] A method and device for predicting flue gas oxygen content load based on joint learning according to an embodiment of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0037] Figure 2 : is a flow chart of a flue gas oxygen content load prediction method based on joint learning provided by an embodiment of the present disclosure. Figure 2 The execution subject is set as the participant, which can be a client or an independent server, and is referred to as the participant here; the central node can be a cloud or an integrated server. Figure 2 As shown, the flue gas oxygen content load prediction method based on joint learning includes:

[0038] S201, the participants respectively determine the sample weights corresponding to the data sets of multiple local devices and the target device data set.

[0039] Specifically, the participants can determine the sample weights corresponding to the data sets of the local multiple groups of devices and the target device data sets respectively in the following manner:

[0040] Step 1: The participant selects the data sets of multiple local devices and the data sets of the target device;

[0041] Among them, the data set can be the steam boiler flue gas temperature, the economizer outlet temperature, the instantaneous value of the flue gas flow, the steam boiler gas temperature, the standard flue gas flow of the steam boiler, the natural gas inlet pressure of the steam boiler, the flue gas flow rate of the steam boiler, the steam boiler condenser inlet flue gas temperature, the steam boiler exhaust temperature, the steam boiler flue gas pressure, the steam boiler condenser inlet pressure, the instantaneous flow rate of the main steam of the steam boiler, the steam boiler operating status, the instantaneous flow rate of the natural gas inlet of the steam boiler, etc. The present invention is not limited to the steam boiler, but can also be a gas furnace or other energy equipment.

[0042] Step 2: merging the data sets of the local multiple groups of devices with the data sets of the target device to obtain merged data;

[0043] Step 3: Using the combined data to train a kernel density estimation model;

[0044] Step 4: According to the kernel density estimation model, the sample weights corresponding to the data sets of the local multiple groups of devices and the data sets of the target devices are determined respectively.

[0045] S202, training a prediction neural network model for multiple groups of devices based on the local data sets of multiple groups of devices and the corresponding sample weights.

[0046] S203, uploading the prediction neural network models of multiple groups of local devices to the central node for model aggregation to obtain an aggregated prediction neural network model.

[0047] Specifically, uploading the prediction neural network models of multiple groups of local devices to the central node for model aggregation to obtain the aggregated prediction neural network model can be achieved in the following way:

[0048] Step 1: Upload the prediction neural network models of multiple local devices to the central node;

[0049] Step 2: respond to the information fed back by the central node;

[0050] Step 3: Receive the aggregated prediction neural network model sent by the central node.

[0051] S204, training the aggregated prediction neural network model according to preset training conditions to obtain a joint prediction model.

[0052] The preset training conditions may include a preset number of training times or an expected convergence state value of the model training, etc.

[0053] Specifically, training the aggregated prediction neural network model according to the preset training conditions to obtain the joint prediction model can be achieved in the following manner:

[0054] Step 1: respond to the aggregated prediction neural network model sent by the central node;

[0055] Step 2: Determine the preset training conditions;

[0056] Step 3: Train the aggregated prediction neural network model according to preset training conditions to obtain a joint prediction model.

[0057] S205, predicting the oxygen content value of the flue gas of the target device according to the joint prediction model and the sample weights corresponding to the target device data set.

[0058] Specifically, according to the joint prediction model and the sample weights corresponding to the target device data set, the prediction of the oxygen content value of the flue gas of the target device can be achieved in the following way:

[0059] Step 1: The participants upload the joint prediction model to the central node for joint learning training;

[0060] Step 2: Joint learning and training of the joint prediction model based on the response and feedback from the intermediate receiving engine;

[0061] Step 3: Send the joint prediction model trained by joint learning to the target device;

[0062] Step 4: Predict the oxygen content value of the flue gas of the target equipment according to the joint prediction model and the sample weight corresponding to the target equipment data set.

[0063] Furthermore, the present invention also provides relevant embodiments for optimizing the prediction neural network model:

[0064] Step 1: The participants use the data sets of multiple local devices to establish training samples for the prediction neural network model;

[0065] Step 2: Use the target device’s data set to build a predictive neural network model test sample;

[0066] Step 3: Obtain sample prediction values ​​based on the prediction neural network model training samples and model test samples;

[0067] Step 4: According to the norm of the error matrix between the sample prediction value and the sample expected value, the fitness value of the prediction neural network model is obtained;

[0068] Step 5: Update the particles in the population of the prediction neural network model according to the fitness value of the prediction neural network model to obtain an optimized prediction neural network model.

[0069] Specifically, it can be further described by the following embodiments:

[0070] The optimized prediction neural network model is as follows:

[0071] (a) Encode the learning parameters of the prediction neural network model to obtain the initial particle population. The encoding rule is: each parameter is represented by a 13-bit binary code, and these parameters are spliced ​​into a particle;

[0072] (b) decoding to obtain parameters of the prediction neural network model, and assigning the parameters to the prediction neural network model;

[0073] (c) using the training samples to train the prediction neural network model;

[0074] (d) using the test sample to test the prediction neural network model and obtain the sample prediction value;

[0075] (e) Selecting the norm of the error matrix between the predicted value and the expected value of the predicted sample as the fitness value;

[0076] (f) Update the particles in the population.

[0077] The parameters involved in particle algorithm update are: speed, position, individual extreme value, and population extreme value. The speed and position update methods are shown in the following formulas (1) and (2). At the same time, in order to prevent blind search of particles, it is generally recommended to limit their position and speed to the interval [-Xmax, Xmax], [-Vmax, Vmax].

[0078]

[0079]

[0080] Variable Description:

[0081] X i =(x i1 ,x i2 ,........x iD ) represents a population of particles with D dimensions, and also represents a solution to the problem

[0082] V i =(v i1 ,v i2 ,.....v iD ) represents the velocity of a population particle with D dimensions

[0083] P i =(p i1 ,p i2 ,.......p iD ) represents the extreme value of individual particles with D dimensions

[0084] P g =(p g1 ,p g2 ,......p gD ) represents the extreme value of the group containing D dimensions

[0085] w is the inertia weight, d=1,2,......D,i=1,2,.......n,k is the current iteration number, Vid is the velocity of the particle, c1, c2 are non-negative constants called acceleration factors, r1, r2 are random numbers distributed in [0,1].

[0086] (g) The optimal crossover is performed on the individuals in the population. The individual particles are updated by crossover with the individual extreme value particles. The crossover method adopts the integer crossover method. First, two crossover positions are selected, and then the individual and the individual extreme value are crossed. The strategy of retaining excellent individuals is adopted for the new individuals. The particles are updated only when the fitness value of the new particle is greater than the fitness value of the old particle.

[0087] (h) Perform optimal crossover on the population. This step is similar to the eighth operation, except that the individual extreme value is replaced by the group extreme value.

[0088] (i) Mutation operation is performed on particles in the population. The mutation operation adopts the method of exchanging two positions within individuals. First, the mutation positions pos1 and pos2 are randomly selected, and then the two mutation positions are exchanged. The strategy of retaining excellent individuals is adopted for the new individuals. The particles are updated only when the fitness value of the new particles is better than that of the old particles.

[0089] (j) Obtain a new population.

[0090] (k) Determine whether the termination condition is met, or whether the maximum number of iterations is reached, or whether the error is less than a specified value.

[0091] (l) If the conditions are not met, go to the third step, otherwise, decode the particle swarm to obtain the initial parameters of the optimal prediction neural network model network.

[0092] A further example of a flue gas oxygen content load prediction method based on joint learning provided by the present invention is as follows: there are data sets of three boilers, namely boiler 1, boiler 2 and boiler 3, among which boiler 1 and boiler 2 are local devices, and boiler 3 is a target device.

[0093] First, data of three boilers, boiler 1, boiler 2, and boiler 3, are selected to obtain local device dataset A and local device dataset B (hereinafter referred to as dataset A and dataset B, respectively), and target device dataset C (hereinafter referred to as dataset C).

[0094] Then, the data from dataset A, dataset B, and dataset C are merged, and a KDE (kernel density estimation model) model is trained using the merged data. The data from dataset A is input into the KDE model to obtain the sample weight of boiler 1 dataset A, and the data from boiler 2 dataset B is input into the KDE model to obtain the sample weight of B.

[0095] Specifically, in a pan-energy station, there will be a lot of boiler data of different models and processes. At this time, the data of multiple boilers can be used to improve the accuracy of the prediction, reduce the installation of boiler sensors, and thus reduce costs and waste of resources.

[0096] Third, at boiler 1, the prediction neural network model can be trained using the data and sample weights of data set A; at boiler 2, the prediction neural network model can be trained using the data and sample weights of data set B. Then, the prediction neural network models trained with data sets A and B are uploaded to the central node for model aggregation. When the central node sends the aggregated model to boilers 1 and 2, boilers 1 and 2 are trained with the aggregated model to obtain a joint prediction model. This is repeated many times until the model is trained to converge, and then the joint prediction model is uploaded to the central node.

[0097] Fourth, the central node sends the joint prediction model trained in the third step to boiler 3, and uses the joint prediction model to predict the oxygen content value of the flue gas at boiler 3.

[0098] According to the technical solution provided by the embodiment of the present disclosure, the participants respectively determine the sample weights corresponding to the data sets of the local multiple groups of equipment and the target equipment data sets; according to the data sets of the local multiple groups of equipment and the corresponding sample weights, the prediction neural network models of the multiple groups of equipment are trained; the prediction neural network models of the local multiple groups of equipment are uploaded to the central node for model aggregation to obtain the aggregated prediction neural network model; the aggregated prediction neural network model is trained according to the preset training conditions to obtain the joint prediction model; according to the joint prediction model and the sample weights corresponding to the target equipment data sets, the smoke oxygen content value of the target equipment is predicted. This solves the problem of resource waste caused by the inability to improve the accuracy of the smoke oxygen content load prediction of energy equipment in the prior art.

[0099] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present application, which will not be described one by one here.

[0100] The following are embodiments of the device disclosed herein, which can be used to execute the method embodiments disclosed herein. For details not disclosed in the device embodiments disclosed herein, please refer to the method embodiments disclosed herein.

[0101] Figure 3 Schematic diagram of a flue gas oxygen content load prediction device based on joint learning provided by an embodiment of the present disclosure. Figure 3 As shown, the flue gas oxygen content load prediction device based on joint learning includes:

[0102] Determination module 301, used by the participants to determine the sample weights corresponding to the data sets of multiple local devices and the target device data set respectively;

[0103] A first training module 302 is used to train a prediction neural network model for multiple groups of devices based on the data sets of multiple groups of local devices and the corresponding sample weights;

[0104] Aggregation module 303, used to upload the prediction neural network models of multiple groups of local devices to the central node for model aggregation to obtain an aggregated prediction neural network model;

[0105] The second training module 304 trains the aggregated prediction neural network model according to preset training conditions to obtain a joint prediction model;

[0106] The prediction module 305 is used to predict the oxygen content value of the smoke of the target device according to the joint prediction model and the sample weight corresponding to the target device data set.

[0107] According to the technical solution provided by the embodiment of the present disclosure, the participants respectively determine the sample weights corresponding to the data sets of the local multiple groups of equipment and the target equipment data sets; according to the data sets of the local multiple groups of equipment and the corresponding sample weights, the prediction neural network models of the multiple groups of equipment are trained; the prediction neural network models of the local multiple groups of equipment are uploaded to the central node for model aggregation to obtain the aggregated prediction neural network model; the aggregated prediction neural network model is trained according to the preset training conditions to obtain the joint prediction model; according to the joint prediction model and the sample weights corresponding to the target equipment data sets, the smoke oxygen content value of the target equipment is predicted. This solves the problem of resource waste caused by the inability to improve the accuracy of the smoke oxygen content load prediction of energy equipment in the prior art.

[0108] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure.

[0109] Figure 4 Schematic diagram of a computer device 4 provided in an embodiment of the present disclosure. Figure 4 As shown, the computer device 4 of this embodiment includes: a processor 401, a memory 402, and a computer program 404 stored in the memory 402 and executable on the processor 401. When the processor 401 executes the computer program 404, the steps in the above-mentioned various method embodiments are implemented. Alternatively, when the processor 401 executes the computer program 404, the functions of each module / unit in the above-mentioned various device embodiments are implemented.

[0110] Exemplarily, the computer program 404 may be divided into one or more modules / units, which are stored in the memory 402 and executed by the processor 401 to complete the present disclosure. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 404 in the computer device 4.

[0111] The computer device 4 may be a desktop computer, a notebook computer, a PDA, a cloud server, or other computer devices. The computer device 4 may include, but is not limited to, a processor 401 and a memory 402. Those skilled in the art will appreciate that Figure 4 It is only an example of computer device 4 and does not constitute a limitation of computer device 4. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0112] The processor 401 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

[0113] The memory 402 may be an internal storage unit of the computer device 4, for example, a hard disk or memory of the computer device 4. The memory 402 may also be an external storage device of the computer device 4, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 4. Further, the memory 402 may also include both an internal storage unit and an external storage device of the computer device 4. The memory 402 is used to store computer programs and other programs and data required by the computer device. The memory 402 may also be used to temporarily store data that has been output or is to be output.

[0114] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0115] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0116] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0117] In the embodiments provided in the present disclosure, it should be understood that the disclosed apparatus / computer equipment and methods can be implemented in other ways. For example, the apparatus / computer equipment embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. Multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection of the apparatus or unit, which may be electrical, mechanical or other forms.

[0118] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0119] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0120] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present disclosure implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, and the computer program code may be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased 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 include electric carrier signals and telecommunication signals.

[0121] The above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should all be included in the protection scope of the present disclosure.

Claims

1. A flue gas oxygen content load prediction method based on joint learning, characterized in that: include: The participants determine the sample weights corresponding to the data sets of the local multiple groups of devices and the data sets of the target devices respectively; According to the data sets of multiple local equipment groups and the corresponding sample weights, a prediction neural network model for multiple equipment groups is trained; Upload the prediction neural network models of multiple groups of local devices to the central node for model aggregation to obtain the aggregated prediction neural network model; Training the aggregated prediction neural network model according to preset training conditions to obtain a joint prediction model; Predict the oxygen content value of the flue gas of the target equipment according to the joint prediction model and the sample weight corresponding to the target equipment data set; The participants determine the sample weights corresponding to the data sets of the local multiple groups of devices and the target device data sets respectively, including: The participant selects the data sets of multiple local devices and the data sets of the target device; Merge the data sets of multiple local devices with the data sets of the target device to obtain merged data; Training a kernel density estimation model using the combined data; According to the kernel density estimation model, sample weights corresponding to the data sets of the local multiple groups of devices and the data sets of the target devices are determined respectively.

2. The method according to claim 1, characterized in that The prediction neural network models of multiple groups of local devices are uploaded to the central node for model aggregation to obtain the aggregated prediction neural network model including: Upload the prediction neural network models of multiple groups of local devices to the central node; Responding to information fed back by the central node; Receive the aggregated prediction neural network model sent by the central node.

3. The method according to claim 1, characterized in that The aggregated prediction neural network model is trained according to the preset training conditions to obtain a joint prediction model including: Responding to the aggregated prediction neural network model sent by the central node; Determine the preset training conditions; The aggregated prediction neural network model is trained according to preset training conditions to obtain a joint prediction model.

4. The method according to claim 1, characterized in that: According to the sample weights corresponding to the joint prediction model and the target equipment data set, the predicted target equipment flue gas oxygen content values ​​include: The participants upload the joint prediction model to the central node for joint learning training; Joint learning training of the joint prediction model based on feedback from the response and collection engines; Sending the joint prediction model trained by joint learning to the target device; According to the joint prediction model and the sample weights corresponding to the target equipment data set, the oxygen content value of the flue gas of the target equipment is predicted.

5. The method according to claim 1, characterized in that Also includes: The participants use the data sets of multiple local devices to establish training samples for the predictive neural network model; Use the target device’s data set to build a predictive neural network model test sample; Obtaining a sample prediction value according to the model training sample and the model testing sample; According to the norm of the error matrix between the sample prediction value and the sample expected value, the fitness value of the prediction neural network model is obtained; The particles in the population in the prediction neural network model are updated according to the fitness value to obtain an optimized prediction neural network model.

6. The method according to any one of claims 1 to 5, characterized in that Both the participants and the central node are in the federated learning architecture.

7. A flue gas oxygen content load prediction device based on joint learning, characterized in that: include: A determination module, used by the participants to determine the sample weights corresponding to the data sets of multiple local devices and the target device data sets respectively; A first training module is used to train a prediction neural network model for multiple groups of devices based on the data sets of multiple groups of local devices and corresponding sample weights; An aggregation module is used to upload the prediction neural network models of multiple groups of local devices to the central node for model aggregation to obtain an aggregated prediction neural network model; The second training module trains the aggregated prediction neural network model according to preset training conditions to obtain a joint prediction model; A prediction module, used to predict the oxygen content value of the flue gas of the target device according to the joint prediction model and the sample weight corresponding to the target device data set; The determination module is specifically used for the participant to select the data sets of multiple local groups of devices and the data sets of the target device; merge the data sets of multiple local groups of devices and the data sets of the target device to obtain merged data; and train the kernel density estimation model using the merged data; According to the kernel density estimation model, sample weights corresponding to the data sets of the local multiple groups of devices and the data sets of the target devices are determined respectively.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images