Multi-target control method and device for energy consumption optimization, electronic equipment and storage medium

Abstract

The invention relates to the field of artificial intelligence, and provides a multi-target control method and device for energy consumption optimization, electronic equipment and a storage medium, and the method comprises the steps: building large-scale and small-scale standard vectors in real time, adaptively adjusting the weights of different optimization targets according to the operation state, solving the problem that a fixed weight cannot adapt to a complex working condition, and improving the optimization efficiency. The energy consumption is remarkably reduced while the temperature control precision is ensured, a hierarchical optimization architecture is provided, the control quality is improved, a candidate scheme is generated by adopting large-scale global optimization, and an optimally executed double-layer decision-making mechanism is determined through small-scale local screening. The method has the beneficial effects that the long-term energy efficiency stability and the short-term dynamic response characteristic are considered, the balance between the global performance and the local performance is realized, the optimization decision is automatically completed in a data driving mode, the comprehensive energy efficiency and the control quality are remarkably improved while the operation safety is guaranteed, and the intelligence and the user experience of the electric fire stove are enhanced.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence, and in particular to a multi-objective control method, apparatus, electronic device, and storage medium for energy consumption optimization. Background Technology

[0002] Traditional electric stove control methods often focus on a single objective, such as constant temperature control or a simple energy-saving mode. They struggle to simultaneously optimize energy consumption, temperature control accuracy, and response speed during dynamic cooking. While existing technologies attempt to introduce multi-objective optimization, they often employ static compromise schemes with fixed weights. These schemes cannot adaptively adjust the optimization focus based on real-time operating conditions (such as cookware material and cooking stage), resulting in limited overall performance. Furthermore, most methods do not consider the differences in optimization objectives over time. Long-term energy consumption and short-term thermal response characteristics require different optimization strategies, and existing solutions lack effective hierarchical coordination mechanisms, making it difficult to achieve a balance between global and local control effects. Summary of the Invention

[0003] The main objective of this invention is to provide a multi-objective control method, device, electronic device, and storage medium for energy consumption optimization, aiming to solve the problem that it is difficult to achieve control effects at both the global and local levels.

[0004] A multi-objective control method for energy consumption optimization, the method comprising: The operating data of the electric stove is collected at a preset sampling period; wherein the operating data includes parameters of multiple preset dimensions; The system extracts operational data within a preset large-scale time window to obtain a large-scale operational state parameter sequence, and extracts operational data within a preset small-scale time window to obtain a small-scale operational state parameter sequence. The large-scale operating state parameter sequence is standardized to obtain a large-scale standardized sequence, and the small-scale operating state parameter sequence is standardized to obtain a small-scale standardized sequence. The large-scale standardized sequence is weighted by a large-scale preset weight vector to obtain a large-scale standard vector, and the small-scale standardized sequence is weighted by a small-scale preset weight vector to obtain a small-scale standard vector. Based on the large-scale standard vector, an optimization problem for multiple first optimization objectives is constructed and solved to obtain a set of multiple intermediate optimization schemes that satisfy the constraints. Based on the small-scale standard vector, the intermediate optimization scheme set is screened a second time to obtain the target optimization scheme, and control commands are generated according to the target optimization scheme to perform closed-loop control of the heating power of the electric stove.

[0005] Further, the step of standardizing the large-scale operating state parameter sequence to obtain a large-scale standardized sequence includes: Obtain multiple preset statistical features of the large-scale operational state parameter sequence; Normalize each of the preset statistical features to obtain the statistical feature value of each preset statistical feature; The statistical feature values ​​are arranged according to a preset statistical feature order to obtain the large-scale standardized sequence.

[0006] Further, the step of standardizing the small-scale operating state parameter sequence to obtain a small-scale standardized sequence includes: The small-scale operating state parameter sequence is encoded using a preset time encoder to obtain time-encoded data, and the small-scale operating state parameter sequence is encoded using a preset state feature encoder to obtain state-encoded data; wherein, the time encoder is used to extract column features from the small-scale operating state parameter sequence for encoding, and the state feature encoder is used to extract lateral features from the small-scale operating state parameter sequence for encoding. The state-coded data and the time-coded data are weighted and fused to obtain a small-scale standardized sequence.

[0007] Furthermore, the step of performing a secondary screening of the intermediate optimization scheme set based on the small-scale standard vector to obtain the target optimization scheme includes: The evaluation value of each scheme in the intermediate optimization scheme set is calculated according to a preset small-scale evaluation function; wherein, the small-scale evaluation function includes at least a temperature tracking error term and a response overshoot term; The scheme with the smallest evaluation value is selected as the target optimization scheme.

[0008] Furthermore, the small-scale evaluation function is expressed as: ; Where TS is the length of the small-scale time window. For the current time, Indicates time as The actual furnace surface temperature Indicates time as The target temperature, where α and β are preset weighting coefficients. Indicates in The maximum difference between the actual furnace surface temperature and the target temperature within the time window.

[0009] Furthermore, the length of the large-scale time window is N times the length of the small-scale time window, where N is an integer greater than or equal to 5.

[0010] Furthermore, the step of constructing and solving optimization problems for multiple first optimization objectives based on the large-scale standard vector to obtain a set of intermediate optimization solutions that satisfy the constraints includes: The large-scale standard vector is decoded into weight coefficients corresponding to the plurality of first optimization objectives to construct a large-scale evaluation function; wherein the plurality of first optimization objectives include at least: minimizing the average energy consumption within a large-scale time window and maximizing the temperature stability within a large-scale time window. Based on the large-scale evaluation function, a Pareto solution set is generated using a multi-objective optimization algorithm within a preset large-scale decision space. Each Pareto solution corresponds to an intermediate optimization scheme that satisfies preset constraints. The preset constraints include at least: maximum output power constraint, device safety temperature constraint, and power change rate smoothing constraint. At least two nondominated solutions are selected from the Pareto solution set to form the intermediate optimization scheme set.

[0011] A multi-objective control device for energy consumption optimization, the device comprising: The data acquisition module is used to acquire the operating data of the electric stove at a preset sampling period; wherein the operating data includes parameters of multiple preset dimensions; The extraction module is used to extract running data within a preset large-scale time window to obtain a large-scale running state parameter sequence, and to extract running data within a preset small-scale time window to obtain a small-scale running state parameter sequence. The standardization module is used to standardize the large-scale operating state parameter sequence to obtain a large-scale standardized sequence, and to standardize the small-scale operating state parameter sequence to obtain a small-scale standardized sequence. The calculation module is used to perform a weighted operation on the large-scale standardized sequence and the large-scale preset weight vector to obtain a large-scale standard vector, and to perform a weighted operation on the small-scale standardized sequence and the small-scale preset weight vector to obtain a small-scale standard vector. The solution module is used to construct and solve optimization problems for multiple first optimization objectives based on the large-scale standard vector, and obtain a set of intermediate optimization solutions that satisfy the constraints. The filtering module is used to perform secondary filtering on the set of intermediate optimization schemes based on the small-scale standard vector to obtain the target optimization scheme, and generate control commands according to the target optimization scheme to perform closed-loop control on the heating power of the electric stove.

[0012] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods described above.

[0013] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any of the methods described above.

[0014] The beneficial effects of this invention are as follows: By constructing large and small-scale standard vectors in real time and adaptively adjusting the weights of different optimization objectives according to the operating status, the problem that fixed weights cannot adapt to complex operating conditions is solved. While ensuring temperature control accuracy, energy consumption is significantly reduced. A hierarchical optimization architecture is proposed, which improves control quality. A two-layer decision-making mechanism is adopted, which generates candidate schemes through large-scale global optimization and determines the optimal execution through small-scale local screening. This takes into account both long-term energy efficiency stability and short-term dynamic response characteristics, achieving a balance between global and local performance. The optimization decision is automatically completed in a data-driven manner, which significantly improves comprehensive energy efficiency and control quality while ensuring operational safety, and enhances the intelligence and user experience of electric stoves.

[0015] From an implementation perspective, the core improvement of this invention lies in the software algorithm design of the control method and system. Its innovation is mainly reflected in the optimization model construction, weight adaptive mechanism, and two-level decision logic, requiring minimal changes to the hardware structure of the electric stove. Existing electric stoves can implement this invention simply by upgrading firmware or configuring standard sensor interfaces, without changing their main heating structure, power module, or basic circuit design. Therefore, it has high compatibility and ease of promotion, significantly lowering the threshold for industrialization. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a multi-objective control method for energy consumption optimization according to an embodiment of the present invention. Figure 2 This is a schematic block diagram of a multi-objective control device for energy consumption optimization according to an embodiment of the present invention; Figure 3 This is a schematic block diagram of the structure of a computer device according to an embodiment of this application.

[0017] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] It should be noted that all directional indicators (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicator will also change accordingly. The connection can be a direct connection or an indirect connection.

[0020] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three kinds of relationships. For example, A and B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0021] Furthermore, in this invention, descriptions involving "first," "second," etc., are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0022] Reference Figure 1 This invention proposes a multi-objective control method for energy consumption optimization, the method comprising: S1: Collect the operating data of the electric stove at a preset sampling period; wherein the operating data includes parameters of multiple preset dimensions; S2: Extract the running data within a preset large-scale time window to obtain a large-scale running state parameter sequence, and extract the running data within a preset small-scale time window to obtain a small-scale running state parameter sequence. S3: Standardize the large-scale operating state parameter sequence to obtain a large-scale standardized sequence, and standardize the small-scale operating state parameter sequence to obtain a small-scale standardized sequence. S4: Perform a weighted operation on the large-scale standardized sequence and the large-scale preset weight vector to obtain a large-scale standard vector, and perform a weighted operation on the small-scale standardized sequence and the small-scale preset weight vector to obtain a small-scale standard vector; S5: Based on the large-scale standard vector, construct and solve optimization problems for multiple first optimization objectives to obtain a set of intermediate optimization schemes that satisfy the constraints. S6: Based on the small-scale standard vector, the intermediate optimization scheme set is screened a second time to obtain the target optimization scheme, and control instructions are generated according to the target optimization scheme to perform closed-loop control on the heating power of the electric stove.

[0023] As described in step S1 above, the operating data of the target electric stove is collected periodically at a preset sampling period. The operating data includes parameters in multiple preset dimensions, such as heating power (P), cooktop temperature (T), input voltage (V), input current (I), and possible characteristics of cooking appliances. These parameters not only reflect the current operating status of the target electric stove but may also be affected by the external environment (such as kitchen temperature and humidity), thus exhibiting high dynamic change characteristics. The preset sampling period should be selected according to the actual operating needs of the target electric stove; for example, it can be set to a few seconds to tens of seconds to capture rapidly changing states. For instance, a shorter sampling period helps to provide timely feedback on current temperature, power, and other information, enabling rapid adjustments in dynamic control. Simultaneously, a filtering mechanism needs to be set to periodically reduce the impact of noise on data acquisition, ensuring the authenticity and reliability of the data. Through continuous sampling, not only can the operating data of the target electric stove be obtained at different time periods, but a time series can also be formed.

[0024] As described in step S2 above, operational data is extracted within a preset large-scale time window to obtain a large-scale operational state parameter sequence, and operational data is extracted within a preset small-scale time window to obtain a small-scale operational state parameter sequence. The large-scale time window is primarily used to analyze long-term trends, typically set to several minutes to several hours, depending on the overall cooking process and the electric stove's operating cycle. Within this window, the system aggregates the collected operational data and calculates various long-term indicators, such as average power consumption and stove surface temperature change trends. This long-term information helps evaluate the electric stove's performance over a period of time, supporting analysis for primary optimization objectives such as energy efficiency and performance stability. In contrast, the small-scale time window focuses on dynamic changes within a shorter timeframe, typically between several seconds and tens of seconds. This window aims to capture the electric stove's rapid response states during real-time operation, such as sudden power changes during rapid heating and peak temperature response. The small-scale state parameter sequence helps the system analyze the electric stove's performance under instantaneous conditions, ensuring efficient energy utilization and rapid response to meet the precision and flexibility requirements during cooking. This dual time window design allows the system to perform data analysis with different focuses in the defined time domain, forming new data structures that support subsequent standardization, optimization analysis, and decision-making, thereby improving the overall control effect.

[0025] As described in step S3 above, the extracted large-scale and small-scale operating state parameter sequences are standardized to obtain large-scale standardized sequences and small-scale standardized sequences. The main purpose of data standardization is to eliminate the dimensional effects of different features and ensure that all parameters can participate in subsequent weight calculations and optimization analyses under the same dimension. A specific standardization method can be the z-score method, which converts the original values ​​into standard normal distribution values ​​by removing the mean and dividing by the standard deviation. This step can be performed using the following formula: z = x − μσ, where x is the original data point, μ is the mean of the batch of data, and σ is the standard deviation of the batch of data. The data z processed in this way becomes a dimensionless standardized value with a more reasonable numerical distribution, making it more suitable for subsequent analysis and decision-making. The large-scale and small-scale standardization processes target different time windows, ensuring that data from different time periods can be smoothly integrated in large-scale analysis and providing a basis for long-term performance evaluation; in small-scale analysis, it improves the comparability and effectiveness of instantaneous performance indicators, making real-time responses under dynamic cooking conditions more accurate. Standardization makes the weights between parameters transparent, ensuring fairness in subsequent calculations and ultimately laying the foundation for building a more scientific optimization model.

[0026] As described in step S4 above, the standardized large-scale and small-scale state parameter sequences are weighted by preset weight vectors to generate large-scale and small-scale standard vectors. The purpose of this step is to introduce corresponding weights to highlight the importance of different parameters to the optimization objective. The construction of the weight vector is based on an understanding of the electric stove's working model. By preset weights, the system can clearly define the priority of different optimization objectives in a multi-objective control context. For example, the weight of heating power can be relatively high to ensure the achievement of energy-saving objectives, while the weight of temperature stability can be appropriately adjusted to ensure the flexibility and responsiveness of the control system. The weight vector should be dynamically adjusted, taking into account user needs, cooking content, and historical operation data. The weighting operation can be implemented using a dot product method, as calculated below: ; ; Wherein, UL and US represent the obtained large-scale and small-scale standard vectors, respectively; ZL and ZS are the standardized large-scale and small-scale state sequences, respectively; WL and WS are pre-defined weight vectors. Through weighted operations, the obtained standard vectors can provide the necessary data structure for subsequent multi-objective optimization construction, ensuring that all considered parameters effectively participate in the comprehensive performance evaluation and decision-making.

[0027] As described in step S5 above, an optimization problem with multiple primary optimization objectives is constructed using large-scale standard vectors, and solved to obtain a set of intermediate optimization solutions that satisfy the constraints. When constructing the optimization problem, multiple primary optimization objectives are first defined, such as minimizing energy consumption, minimizing temperature fluctuations, and maximizing user experience. Each optimization objective is closely related to specific user needs, the cooking environment, and the working mechanism of the electric stove. Through effective trade-offs and optimization, the best comprehensive solution is sought. During the solution process, multi-objective optimization algorithms (such as particle swarm optimization, genetic algorithms, or simulated annealing algorithms) can be applied. These algorithms can simultaneously consider multiple objectives and generate effective candidate solutions. By running these algorithms within a preset decision space, the system will quickly approach the Pareto front of each objective, forming a set of non-dominated solutions, i.e., a set of intermediate optimization solutions. Each intermediate solution is the result produced for the optimization objective under certain constraints (such as maximum power limits and temperature safety limits). By obtaining multiple intermediate optimization solutions, the system compares the advantages and disadvantages of different solutions, providing rich selection criteria for the next step of small-scale optimization screening, ensuring that the most suitable control strategy for the current cooking conditions can be selected in actual operation. This process not only improves the flexibility of electric stoves, but also achieves a dual improvement in technology and user experience.

[0028] As described in step S6 above, based on the small-scale standard vector, the set of intermediate optimization schemes is subjected to secondary screening to obtain the target optimization scheme. Control commands are then generated according to the target optimization scheme to perform closed-loop control of the electric stove's heating power. The secondary screening process includes a detailed evaluation of the performance of various intermediate optimization schemes, focusing on instantaneous performance within a small-scale time window. Specific evaluation criteria may include predicted temperature tracking error, response overshoot, and adjustment time, which effectively reflect the electric stove's performance in a dynamic operating environment. Simultaneously, the purpose of secondary screening is to ensure that the selected target optimization scheme performs optimally on a small scale, fully meeting the user's immediate response requirements for heating power. When generating control commands, the system considers the screened target scheme and converts it into specific control parameters, such as power settings and heating cycles. These control commands are sent to the electric stove's control system to achieve closed-loop control, ensuring real-time adjustment of temperature and power, and maintaining the electric stove in the desired operating state during cooking. This ensures that the system not only performs excellently in long-term thermal stability but also possesses strong dynamic response capabilities, adapting to ever-changing cooking needs and effectively improving the user experience. Through closed-loop control, electric stoves can achieve more precise heating effects, continuously optimize their energy consumption, and ultimately achieve a synergistic optimization effect for multiple objectives.

[0029] In one embodiment, step S3, which standardizes the large-scale operational state parameter sequence to obtain a large-scale standardized sequence, includes: S301: Obtain multiple preset statistical features of the large-scale operating state parameter sequence; S302: Normalize each of the preset statistical features to obtain the statistical feature value of each preset statistical feature; S303: Arrange the statistical feature values ​​according to the preset statistical feature order to obtain the large-scale standardized sequence.

[0030] As described in step S301 above, multiple preset statistical features of the large-scale operating state parameter sequence are obtained. These preset statistical features are extracted by analyzing the large-scale operating state parameter sequence. The preset statistical features may include multiple features such as mean, variance, maximum value, minimum value, skewness, and kurtosis, which can effectively describe the distribution and characteristics of the data. Typically, the mean reflects the central position of the data, embodying the average working state of the electric stove within a given time window; variance measures the degree of data fluctuation, reflecting the stability of the operating state; the maximum and minimum values ​​capture the system's extreme behavior within this time period, indicating possible abnormal states or extreme temperature values; skewness and kurtosis provide more detailed distribution characteristics, revealing the asymmetry and peak degree of the data distribution. Through the extraction of these statistical features, the system can better understand the operating trends and potential problems during operation, providing reliable data support for standardization processing and subsequent optimization decisions. This stage of data processing lays a solid foundation for the scientific validity and accuracy of the method, while ensuring the reliability and practicality of the information contained in the final generated large-scale standardized sequence.

[0031] As described in step S302 above, each of the preset statistical features is normalized to obtain the statistical feature value of each preset statistical feature. The extracted preset statistical features are then normalized to ensure they can be compared and analyzed under the same dimension. Normalization is a crucial step in data preprocessing. By uniformly transforming data at different scales to the same standard, the impact of differences in dimensions on data analysis can be reduced. Generally, there are various normalization methods, among which linear normalization and Z-score normalization are more common. Linear normalization scales the value of each statistical feature to the [0,1] interval, while Z-score normalization adjusts the feature to a standard normal distribution by subtracting the mean and dividing by the standard deviation. For statistical feature values, this process clearly transforms the values ​​into a dimensionless form, facilitating the subsequent synthesis of standardized sequences. After normalization, each statistical feature value can be effectively compared within the same range, providing a foundation for constructing standardized sequences.

[0032] As described in step S303 above, the normalized statistical feature values ​​are arranged in a preset order to form the final large-scale standardized sequence. This step emphasizes the organization and structured presentation of information in the sequence, making subsequent parameter analysis and comparison more systematic and organized. The preset order of statistical features usually follows specific logical relationships or priorities, such as being determined based on the importance of each feature in the performance optimization of the electric stove. This order not only helps to clarify the position of each statistical feature in behavioral analysis but also provides the direct input order required by subsequent optimization algorithms, making the algorithm processing more efficient.

[0033] In one embodiment, step S3, which standardizes the small-scale operating state parameter sequence to obtain a small-scale standardized sequence, includes: S311: The small-scale operating state parameter sequence is encoded using a preset time encoder to obtain time-coded data, and the small-scale operating state parameter sequence is encoded using a preset state feature encoder to obtain state-coded data; wherein, the time encoder is used to extract column features from the small-scale operating state parameter sequence for encoding, and the state feature encoder is used to extract lateral features from the small-scale operating state parameter sequence for encoding. S312: Perform weighted fusion processing on the state-coded data and the time-coded data to obtain a small-scale standardized sequence.

[0034] As described in step S311 above, a preset time encoder is used to encode the small-scale operating state parameter sequence to extract its columnar features and obtain the corresponding time-coded data; a state feature encoder is used to encode the same parameter sequence to capture its lateral features and obtain state-coded data. This process aims to transform the original operating state parameter sequence into coded data with higher information density and interpretability to adapt to subsequent standardization processing. Specifically, the time encoder focuses on extracting time-related dynamic change features from the small-scale operating state parameter sequence. For example, within a specified small-scale time window, the temperature change of an electric stove may manifest as the rate of temperature rise, stability, and instantaneous power at a specific moment, all of which are columnar features proportional to time. Through the time encoder, this information can be compressed and transformed into vector form. At the same time, the state feature encoder focuses on extracting features related to the state of the electric stove, such as the current heating power level, the operating status of the equipment (e.g., start / stop status), and other features related to the adjustment of the equipment's operation. This state information is usually lateral, reflecting a snapshot of the equipment's operation at a specific moment, which can help the system better understand its current state. Therefore, by combining these two encoders, we can comprehensively capture the important features in the small-scale operating state parameter sequence, making subsequent analysis more effective and accurate.

[0035] As described in step S312 above, the encoded state-coded data and time-coded data are weighted and fused to generate a small-scale standardized sequence. The specific implementation of the weighted fusion can employ a simple linear weighting method or a more complex nonlinear weighting model, depending on the system's design goals and implementation requirements. The linear weighting method typically involves multiplying each encoded data point by one using a preset weight vector and then summing the results, for example: ; Where Zsmall is the final generated small-scale normalized sequence, S is the state-coded data, and T is the time-coded data. and These are preset weighting coefficients, which allow for flexible adjustment of the importance of different features in the final output. Through weighted fusion, the system can organically integrate state and time data, forming a unified standardized sequence that includes dynamic changes over time and current state information. This fusion not only helps improve the accuracy of subsequent analysis but also enhances the effectiveness of the electric stove's control strategy in short-term dynamic response, enabling the optimization algorithm to make more accurate decisions when processing real-time data. Ultimately, the small-scale standardized sequence provides essential high-quality input for the closed-loop control of the entire system, ensuring the electric stove can respond quickly in instantaneous states, significantly improving the user experience and meeting real-time cooking needs.

[0036] In one embodiment, step S6, which involves performing a secondary screening of the intermediate optimization scheme set based on the small-scale standard vector to obtain the target optimization scheme, includes: S601: Calculate the evaluation value of each scheme in the intermediate optimization scheme set according to the preset small-scale evaluation function; wherein, the small-scale evaluation function includes at least a temperature tracking error term and a response overshoot term; S602: Select the scheme with the smallest evaluation value as the target optimization scheme.

[0037] As described in steps S601-S602 above, the evaluation value of each scheme in the set of intermediate optimization schemes is calculated according to the preset small-scale evaluation function. Through the design of the evaluation function, the performance of each scheme within a small-scale time window is quantified to select the optimal scheme. The small-scale evaluation function may include a temperature tracking error term and a response overshoot term. The temperature tracking error term measures the difference between the stove surface temperature and the set target temperature when executing a specific intermediate scheme. This indicator can be represented by the root mean square error (RMSE). The response overshoot term indicates the degree to which the stove surface temperature exceeds the set target temperature in the initial response when the electric stove adjusts its heating power. This indicator is crucial for understanding the transient performance of the system; a larger overshoot indicates poorer system stability during the response process, which may lead to larger temperature fluctuations when cooking food, affecting the final cooking effect. After combining these indicators, the system can generate a comprehensive evaluation value for each intermediate scheme, i.e., a small-scale evaluation value.

[0038] Based on the previously calculated evaluation values, the scheme with the smallest evaluation value is selected as the target optimization scheme. This process is a crucial step in the secondary screening, its core purpose being to determine the intermediate optimization scheme that performs best in terms of small-scale dynamic response by comparing the magnitudes of the evaluation values. Minimizing the evaluation value means that the selected scheme can provide the best temperature tracking accuracy and response robustness within a small-scale time window. Selecting the scheme that achieves the lowest temperature tracking error and the lowest response overshoot can effectively improve user satisfaction during the cooking process. The final target optimization scheme will become the execution scheme of the electric stove control system. The control commands generated by this scheme can be used to dynamically adjust the heating power to meet real-time cooking needs.

[0039] In one embodiment, the small-scale evaluation function is expressed as: ; Where TS is the length of the small-scale time window. For the current time, Indicates time as The actual furnace surface temperature Indicates time as The target temperature, where α and β are preset weighting coefficients. Indicates in The maximum difference between the actual furnace surface temperature and the target temperature within the time window.

[0040] In this embodiment, the first part of the formula integrates the actual stove surface temperature and the corresponding target temperature within a small-scale time window. Through integration, not only the instantaneous value of the temperature tracking error can be obtained, but the dynamic trend of temperature change throughout the entire time window can also be captured. This means that the small-scale evaluation function not only considers the temperature state at a single point in time, but also integrates all temperature data within the time window, thus forming a more comprehensive performance indicator. For example, integration can reflect how the temperature changes over a certain period (such as from the start of heating to stable heating), allowing the evaluation to focus not only on extreme instantaneous states but also on long-term temperature performance. This comprehensive measurement method can effectively reflect the actual operating status of the electric stove during dynamic operation, helping the optimization algorithm make more rational decisions. The second part considers the maximum temperature within the time window, which effectively reflects the temperature performance at peak times, thereby improving control stability. When the actual stove surface temperature is high at certain moments, it reflects the cooling and heating status, thus promoting dynamic adjustment strategies. This mechanism ensures that the system will not ignore adjustments when encountering sudden overheating, avoiding temperature control failure and ensuring safety and reliability during use. Integrating the actual cooktop temperature and the corresponding target temperature within a small-scale time window, and then weighting the sum with the maximum value, significantly improves the comprehensiveness of the analytical data while enhancing the stability of temperature control and the system's dynamic response capability. This design not only improves the performance of the electric stove but also ensures a satisfactory user experience during cooking, making it an important means of achieving efficient, multi-objective optimization.

[0041] In one embodiment, the length of the large-scale time window is N times the length of the small-scale time window, where N is an integer greater than or equal to 5.

[0042] In this embodiment, firstly, setting the large-scale window to more than five times the small-scale window ensures that when analyzing and optimizing the performance of the electric stove, consideration of its long-term operating status is significantly better than short-term fluctuations. Long-term data analysis allows the system to capture the overall trend of equipment performance, such as energy consumption reduction and temperature variation range, thereby gaining a more comprehensive understanding of the electric stove's performance under different working states and environmental conditions. Secondly, a longer large-scale time window helps mitigate the impact of short-term noise on the analysis results. Within a small-scale time window, data may be significantly affected by instantaneous conditions, external interference, or occasional equipment failures. When the small-scale window is short, these instantaneous fluctuations may lead to unstable output, making it difficult to reflect the true performance of the equipment. By setting a relatively long large-scale time window, the representativeness of the data can be enhanced statistically, making the evaluation more accurate and reliable. In cooking practice, users want to quickly reach the preset temperature and maintain a stable temperature without frequent fluctuations. Setting a large-scale time window greater than five times the small-scale window allows for long-term tracking of temperature changes, reducing frequent and large-amplitude temperature fluctuations caused by short-term adjustments, thereby improving the user's cooking experience. In summary, setting the large-scale time window to be more than five times the length of the small-scale time window not only optimizes the stability and reliability of the data analysis and decision-making models, but also effectively stratifies the control strategy, enhances dynamic adjustment capabilities, and thus improves the overall performance of the electric stove and the user experience. This design choice is a powerful response to practical application needs, ensuring that the control system maintains superior performance in complex and ever-changing cooking environments.

[0043] In one embodiment, step S5, which involves constructing and solving optimization problems for multiple first optimization objectives based on the large-scale standard vector to obtain a set of intermediate optimization solutions that satisfy the constraints, includes: S501: Decode the large-scale standard vector into weight coefficients corresponding to the plurality of first optimization objectives to construct a large-scale evaluation function; wherein, the plurality of first optimization objectives include at least: minimizing the average energy consumption within the large-scale time window and maximizing the temperature stability within the large-scale time window; S502: Based on the large-scale evaluation function, a Pareto solution set is generated using a multi-objective optimization algorithm within a preset large-scale decision space, wherein each Pareto solution corresponds to an intermediate optimization scheme that satisfies preset constraints; wherein the preset constraints include at least: maximum output power constraint, device safety temperature constraint, and power change rate smoothing constraint. S503: Select at least two non-dominated solutions from the Pareto solution set to form the intermediate optimization scheme set.

[0044] As described in step S501 above, the large-scale standard vector is decoded into weight coefficients corresponding to multiple first optimization objectives to construct a large-scale evaluation function. The large-scale standard vector is obtained by standardizing and weighting the large-scale operating state parameters. This vector encodes multiple key parameters related to the operation of the electric stove. Through the decoding process, the system extracts multiple weight coefficients, which represent the priority of each optimization objective in the current operating state. For example, minimizing average energy consumption may be considered more important than temperature stability in daily use, so the corresponding weight coefficient may be set higher. These weights can also be flexibly adjusted based on historical operating data or user preferences to reflect the actual needs in different cooking scenarios. The constructed large-scale evaluation function typically takes the form of a linear combination, which can integrate the weights of various optimization objectives. For example, the evaluation function can be expressed as: In this equation, w3 and w4 are weighting coefficients related to average energy consumption and temperature stability, respectively; E is the calculated energy consumption index; and S is the evaluation index for temperature stability. Through this evaluation function, the system can quickly assess the benefits of different technical solutions during subsequent optimization, thereby selecting the optimal solution.

[0045] As described in step S502 above, based on the large-scale evaluation function, a Pareto solution set is generated within a preset large-scale decision space using a multi-objective optimization algorithm. Each Pareto solution corresponds to an intermediate optimization scheme that satisfies preset constraints. These preset constraints include at least: maximum output power constraints, device safety temperature constraints, and power change rate smoothing constraints. Specifically, a multi-objective optimization algorithm is applied within the preset large-scale decision space to generate the Pareto solution set. Pareto optimization seeks a set of solutions where each solution is a non-dominated solution, meaning no other solution performs better across all objectives. This method is suitable for handling multiple, potentially conflicting optimization objectives, such as in this invention, where both energy minimization and temperature stability need to be considered. Through multi-objective optimization algorithms (such as genetic algorithms, particle swarm optimization, etc.), the system can search the obtained solution space to obtain multiple feasible solutions. During the generation process, the system also needs to meet a series of preset constraints, such as: maximum output power constraint: ensuring that the equipment does not exceed the predetermined maximum power during operation to prevent overload; this can be obtained from the electrical specifications and equipment design of the electric stove. Device safety temperature constraint: ensuring that the furnace surface temperature does not exceed the set maximum equipment safety threshold to ensure safe use; this can be obtained based on material properties, actual equipment measurement data, and industry standards. Power change rate smoothing constraint: limiting the rate of change of power settings to avoid excessive instantaneous power fluctuations and protect the stability of system operation; this can be obtained based on the control characteristics of the equipment and user experience. By comprehensively considering these constraints, the system can ensure that the final Pareto solution set is not only optimal in terms of the objective but also feasible in actual operation.

[0046] As described in step S503 above, at least two non-dominated solutions are selected from the generated Pareto solution set to form an intermediate optimization scheme set. Each solution in the Pareto solution set represents a trade-off between various optimization objectives, and the non-dominated solutions are its core feature, indicating that these schemes exhibit superior performance on certain optimization objectives. The selection criteria can be based on the differences in the performance of these non-dominated solutions on the objectives and their adaptability to operating conditions, ensuring that the resulting scheme set can cover a variety of effective choices, such as one scheme focusing on energy efficiency optimization and another focusing on temperature stability. This diversity of choices can provide more options for subsequent small-scale secondary screening steps, effectively improving the system's adaptability. In addition, selecting multiple non-dominated solutions is beneficial to users, allowing them to switch between different operating strategies at any time based on their personal cooking needs or preferences, thereby improving the user experience. The schemes selected from the intermediate scheme set will be directly used in subsequent steps, ensuring that the superiority and flexibility of the electric stove are effectively combined, achieving optimization effects while meeting different user needs.

[0047] Reference Figure 2 The present invention also provides a multi-objective control device for energy consumption optimization, the device comprising: The acquisition module 902 is used to acquire the operating data of the electric stove at a preset sampling period; wherein, the operating data includes parameters of multiple preset dimensions; The extraction module 904 is used to extract running data within a preset large-scale time window to obtain a large-scale running state parameter sequence, and to extract running data within a preset small-scale time window to obtain a small-scale running state parameter sequence. The standardization module 906 is used to standardize the large-scale operating state parameter sequence to obtain a large-scale standardized sequence, and to standardize the small-scale operating state parameter sequence to obtain a small-scale standardized sequence. The calculation module 908 is used to perform a weighted operation on the large-scale standardized sequence and the large-scale preset weight vector to obtain a large-scale standard vector, and to perform a weighted operation on the small-scale standardized sequence and the small-scale preset weight vector to obtain a small-scale standard vector. The solution module 910 is used to construct and solve optimization problems for multiple first optimization objectives based on the large-scale standard vector, and obtain a set of multiple intermediate optimization schemes that satisfy the constraints. The filtering module 912 is used to perform secondary filtering on the set of intermediate optimization schemes based on the small-scale standard vector to obtain the target optimization scheme, and generate control instructions according to the target optimization scheme to perform closed-loop control on the heating power of the electric stove.

[0048] In one embodiment, the standardization module 906 includes: A preset statistical feature acquisition submodule is used to acquire multiple preset statistical features of the large-scale operating state parameter sequence; The normalization processing submodule is used to normalize each of the preset statistical features to obtain the statistical feature value of each preset statistical feature; The permutation submodule is used to arrange the statistical feature values ​​according to a preset statistical feature order to obtain the large-scale standardized sequence.

[0049] In one embodiment, the standardization module 906 includes: The encoding submodule is used to encode the small-scale operating state parameter sequence using a preset time encoder to obtain time-encoded data, and to encode the small-scale operating state parameter sequence using a preset state feature encoder to obtain state-encoded data; wherein, the time encoder is used to extract column features from the small-scale operating state parameter sequence for encoding, and the state feature encoder is used to extract lateral features from the small-scale operating state parameter sequence for encoding. The weighted fusion processing submodule is used to perform weighted fusion processing on the state-coded data and the time-coded data to obtain a small-scale standardized sequence.

[0050] In one embodiment, the filtering module 912 includes: The evaluation value calculation submodule is used to calculate the evaluation value of each scheme in the intermediate optimization scheme set according to a preset small-scale evaluation function; wherein, the small-scale evaluation function includes at least a temperature tracking error term and a response overshoot term; The target optimization scheme selection submodule is used to select the scheme with the smallest evaluation value as the target optimization scheme.

[0051] In one embodiment, the small-scale evaluation function is expressed as: ; Where TS is the length of the small-scale time window. For the current time, Indicates time as The actual furnace surface temperature Indicates time as The target temperature, where α and β are preset weighting coefficients. Indicates in The maximum difference between the actual furnace surface temperature and the target temperature within the time window.

[0052] In one embodiment, the length of the large-scale time window is N times the length of the small-scale time window, where N is an integer greater than or equal to 5.

[0053] In one embodiment, the solving module 910 includes: The decoding submodule is used to decode the large-scale standard vector into weight coefficients corresponding to the plurality of first optimization objectives, so as to construct a large-scale evaluation function; wherein the plurality of first optimization objectives include at least: minimizing the average energy consumption within a large-scale time window and maximizing the temperature stability within a large-scale time window. The Pareto solution set generation submodule is used to generate a Pareto solution set based on the large-scale evaluation function within a preset large-scale decision space using a multi-objective optimization algorithm. Each Pareto solution corresponds to an intermediate optimization scheme that satisfies preset constraints. The preset constraints include at least: maximum output power constraint, device safety temperature constraint, and power change rate smoothing constraint. The non-dominated solution selection submodule is used to select at least two non-dominated solutions from the Pareto solution set to form the intermediate optimization scheme set.

[0054] The beneficial effects of this invention are as follows: By constructing large and small-scale standard vectors in real time and adaptively adjusting the weights of different optimization objectives according to the operating status, the problem that fixed weights cannot adapt to complex operating conditions is solved. While ensuring temperature control accuracy, energy consumption is significantly reduced. A hierarchical optimization architecture is proposed, which improves control quality. A two-layer decision-making mechanism is adopted, which generates candidate schemes through large-scale global optimization and determines the optimal execution through small-scale local screening. This takes into account both long-term energy efficiency stability and short-term dynamic response characteristics, achieving a balance between global and local performance. The optimization decision is automatically completed in a data-driven manner, which significantly improves comprehensive energy efficiency and control quality while ensuring operational safety, and enhances the intelligence and user experience of electric stoves.

[0055] Reference Figure 3 This application also provides a computer device, which may be a server, and its internal structure may be as follows: Figure 3 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores various operational data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it can implement the energy-optimized multi-objective control method described in any of the above embodiments.

[0056] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer equipment on which the present application is applied.

[0057] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, can implement the energy consumption optimization multi-objective control method described in any of the above embodiments.

[0058] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media provided in this application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0059] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0060] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0061] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0062] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A multi-objective control method for energy consumption optimization, characterized in that, The method includes: The operating data of the electric stove is collected at a preset sampling period; wherein the operating data includes parameters of multiple preset dimensions; The system extracts operational data within a preset large-scale time window to obtain a large-scale operational state parameter sequence, and extracts operational data within a preset small-scale time window to obtain a small-scale operational state parameter sequence. The large-scale operating state parameter sequence is standardized to obtain a large-scale standardized sequence, and the small-scale operating state parameter sequence is standardized to obtain a small-scale standardized sequence. The large-scale standardized sequence is weighted by a large-scale preset weight vector to obtain a large-scale standard vector, and the small-scale standardized sequence is weighted by a small-scale preset weight vector to obtain a small-scale standard vector. Based on the large-scale standard vector, an optimization problem for multiple first optimization objectives is constructed and solved to obtain a set of multiple intermediate optimization schemes that satisfy the constraints. Based on the small-scale standard vector, the intermediate optimization scheme set is screened a second time to obtain the target optimization scheme, and control commands are generated according to the target optimization scheme to perform closed-loop control of the heating power of the electric stove.

2. The multi-objective control method for energy consumption optimization according to claim 1, characterized in that, The step of standardizing the large-scale operational state parameter sequence to obtain a large-scale standardized sequence includes: Obtain multiple preset statistical features of the large-scale operational state parameter sequence; Normalize each of the preset statistical features to obtain the statistical feature value of each preset statistical feature; The statistical feature values ​​are arranged according to a preset statistical feature order to obtain the large-scale standardized sequence.

3. The multi-objective control method for energy consumption optimization according to claim 1, characterized in that, The step of standardizing the small-scale operating state parameter sequence to obtain a small-scale standardized sequence includes: The small-scale operating state parameter sequence is encoded using a preset time encoder to obtain time-encoded data, and the small-scale operating state parameter sequence is encoded using a preset state feature encoder to obtain state-encoded data; wherein, the time encoder is used to extract column features from the small-scale operating state parameter sequence for encoding, and the state feature encoder is used to extract lateral features from the small-scale operating state parameter sequence for encoding. The state-coded data and the time-coded data are weighted and fused to obtain a small-scale standardized sequence.

4. The multi-objective control method for energy consumption optimization according to claim 1, characterized in that, The step of performing a secondary screening of the intermediate optimization scheme set based on the small-scale standard vector to obtain the target optimization scheme includes: The evaluation value of each scheme in the intermediate optimization scheme set is calculated according to a preset small-scale evaluation function; wherein, the small-scale evaluation function includes at least a temperature tracking error term and a response overshoot term; The scheme with the smallest evaluation value is selected as the target optimization scheme.

5. The multi-objective control method for energy consumption optimization according to claim 4, characterized in that, The small-scale evaluation function is expressed as follows: ； Where TS is the length of the small-scale time window. For the current time, Indicates time as The actual furnace surface temperature Indicates time as The target temperature, where α and β are preset weighting coefficients. Indicates in The maximum difference between the actual furnace surface temperature and the target temperature within the time window.

6. The multi-objective control method for energy consumption optimization according to claim 1, characterized in that, The length of the large-scale time window is N times the length of the small-scale time window, where N is an integer greater than or equal to 5.

7. The multi-objective control method for energy consumption optimization according to claim 1, characterized in that, The step of constructing and solving optimization problems for multiple first optimization objectives based on the large-scale standard vector to obtain a set of intermediate optimization solutions that satisfy the constraints includes: The large-scale standard vector is decoded into weight coefficients corresponding to the plurality of first optimization objectives to construct a large-scale evaluation function; wherein the plurality of first optimization objectives include at least: minimizing the average energy consumption within a large-scale time window and maximizing the temperature stability within a large-scale time window. Based on the large-scale evaluation function, a Pareto solution set is generated using a multi-objective optimization algorithm within a preset large-scale decision space. Each Pareto solution corresponds to an intermediate optimization scheme that satisfies preset constraints. The preset constraints include at least: maximum output power constraint, device safety temperature constraint, and power change rate smoothing constraint. At least two nondominated solutions are selected from the Pareto solution set to form the intermediate optimization scheme set.

8. A multi-objective control device for energy consumption optimization, characterized in that, The device includes: The data acquisition module is used to acquire the operating data of the electric stove at a preset sampling period; wherein the operating data includes parameters of multiple preset dimensions; The extraction module is used to extract running data within a preset large-scale time window to obtain a large-scale running state parameter sequence, and to extract running data within a preset small-scale time window to obtain a small-scale running state parameter sequence. The standardization module is used to standardize the large-scale operating state parameter sequence to obtain a large-scale standardized sequence, and to standardize the small-scale operating state parameter sequence to obtain a small-scale standardized sequence. The calculation module is used to perform a weighted operation on the large-scale standardized sequence and the large-scale preset weight vector to obtain a large-scale standard vector, and to perform a weighted operation on the small-scale standardized sequence and the small-scale preset weight vector to obtain a small-scale standard vector. The solution module is used to construct and solve optimization problems for multiple first optimization objectives based on the large-scale standard vector, and obtain a set of multiple intermediate optimization solutions that satisfy the constraints. The filtering module is used to perform secondary filtering on the set of intermediate optimization schemes based on the small-scale standard vector to obtain the target optimization scheme, and generate control commands according to the target optimization scheme to perform closed-loop control on the heating power of the electric stove.

9. A computer-readable storage medium, characterized in that, The system contains a computer program that, when executed by a processor, causes the processor to perform the steps of the energy consumption optimization multi-objective control method as described in any one of claims 1 to 7.

10. An electronic device, characterized in that, The device includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the energy consumption optimization multi-objective control method as described in any one of claims 1 to 7.

Images