Intelligent temperature adjusting method and system of temperature controller

Abstract

The invention provides an intelligent temperature adjustment method and system for a temperature controller, and relates to the technical field of temperature control, and the method comprises the steps: obtaining an original multi-source temperature control data set of a target controlled object; based on a neighborhood collaborative voting mechanism, performing credible processing on the original multi-source temperature control data set to generate a credible temperature field; calculating a temperature credible weight matrix according to the credible temperature field and the original temperature field; performing time sequence alignment and feature fusion processing on the credible temperature field and the temperature credible weight matrix, and determining an input sequence; performing multi-step temperature prediction processing on the input sequence through a time sequence prediction model based on thermodynamic constraint, and outputting a temperature prediction sequence in a future control period; according to the temperature prediction sequence, a rapid analysis prediction control model is constructed; solving the rapid analysis prediction control model, and determining an optimal control instruction in the current control period; and executing the optimal control instruction, and obtaining temperature feedback data after execution. According to the invention, the temperature tracking precision can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of temperature control, in particular to a temperature controller intelligent temperature adjustment method and system. BACKGROUND

[0002] With the continuous improvement of building intelligence level and HVAC (Heating, Ventilation and Air Conditioning) automation degree, temperature controllers have been widely used in various scenes such as residences, office buildings and industrial sites, for realizing automatic adjustment and energy consumption management of indoor environment temperature. In actual operation, the temperature controller usually relies on multiple temperature sensors to collect indoor and outdoor environment information, and outputs corresponding temperature adjustment instructions according to the control strategy.

[0003] For the temperature control problem, it is of great significance to introduce an intelligent temperature adjustment method. By predicting and controlling the temperature controller, the energy consumption can be effectively reduced while meeting the comfort requirements, and the operation efficiency and response ability of the temperature control system can be improved. Especially under complex environment and multiple disturbance conditions, more stable, accurate and energy-saving temperature adjustment effect can be realized, which has significant engineering application value and popularization significance.

[0004] However, the existing adjustment method still has certain limitations. On the one hand, some methods excessively rely on single sensor data or simple data fusion strategy, lack effective identification and credibility evaluation of abnormal temperature data, and are easy to cause control decision to be affected by noise or local anomaly; on the other hand, the existing methods mostly use empirical model or pure data-driven prediction mode, ignoring the thermodynamic constraints followed by temperature change, resulting in insufficient stability and reliability of prediction results under complex working conditions, and further affecting the control effect. SUMMARY

[0005] In order to solve the technical problems that the existing adjustment method excessively relies on single sensor data or simple data fusion strategy, lacks effective identification and credibility evaluation of abnormal temperature data, is easy to cause control decision to be affected by noise or local anomaly, and further, the existing method mostly uses empirical model or pure data-driven prediction mode, ignores the thermodynamic constraints followed by temperature change, resulting in insufficient stability and reliability of prediction results under complex working conditions, and further affecting the control effect, the present application provides a temperature controller intelligent temperature adjustment method and system.

[0006] The technical scheme provided by the embodiments of the present application is as follows: The first aspect of the embodiments of the present application provides a temperature controller intelligent temperature adjustment method, comprising: S1: obtaining an original multi-source temperature control data set of a target controlled object; S2: based on a neighborhood collaborative voting mechanism, performing credibility processing on the original multi-source temperature control data set to generate a credible temperature field; S3: calculating a temperature credibility weight matrix according to the credible temperature field and the original temperature field; S4: performing time sequence alignment and feature fusion processing on the credible temperature field and the temperature credibility weight matrix to determine an input sequence; S5: performing multi-step temperature prediction processing on the input sequence through a time sequence prediction model based on thermodynamic constraints to output a temperature prediction sequence in a future control period; S6: constructing a fast analytical predictive control model according to the temperature prediction sequence; S7: solving the fast analytical predictive control model to determine an optimal control instruction in the current control period; S8: executing the optimal control instruction and obtaining temperature feedback data after execution.

[0007] A second aspect of the embodiment of the present application provides a temperature controller intelligent temperature regulating system, comprising: a processor; a memory, the memory storing computer readable instructions, the computer readable instructions being executed by the processor to implement the temperature controller intelligent temperature regulating method of the first aspect.

[0008] A third aspect of the embodiment of the present application provides a computer readable storage medium, which stores a computer program, the program being executed by a processor to implement the temperature controller intelligent temperature regulating method of the first aspect.

[0009] The technical scheme provided by the embodiment of the present application has at least the following beneficial effects: In the embodiment of the present application, by introducing multi-source temperature data credibility processing based on a neighborhood collaborative voting mechanism, a credible temperature field is constructed and a temperature credibility weight matrix is generated, which effectively suppresses the influence of abnormal sensor data and environmental noise on temperature perception results. At the same time, the credible temperature information is time sequence fused with control history and outdoor environment data, and thermodynamic constraints are introduced in the temperature prediction process to realize high-precision, multi-step prediction of indoor temperature changes in the future control period. On this basis, by constructing a fast analytical predictive control model and solving the optimal control instruction, the temperature regulation process realizes the improvement of temperature tracking precision, the more stable control process, and the significant improvement of system robustness and energy efficiency level under the premise of meeting physical constraints and system operation safety requirements, thereby improving the intelligent temperature regulating performance and engineering application reliability of the temperature controller under complex working conditions. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced in the following. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A flowchart of a temperature controller intelligent temperature regulating method provided by an embodiment of the present application.

[0012] Figure 2 A structural diagram of a temperature controller intelligent temperature regulating system provided by an embodiment of the present application. DETAILED DESCRIPTION

[0013] The technical solutions in the present application will be described below with reference to the drawings.

[0014] In the embodiments of the present application, the words such as “example”, “for example” are used to represent an example, illustration or description. Any embodiment or design scheme described as “example” in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. In fact, the word “example” is intended to present the concept in a specific way. In addition, in the embodiments of the present application, the meaning expressed by “and / or” can be both, or can be one of the two.

[0015] In the embodiments of the present application, “image” and “picture” can be used interchangeably at times. It should be pointed out that when the distinction is not emphasized, the meanings expressed are consistent. “Of”, “corresponding” and “corresponding” can be used interchangeably at times. It should be pointed out that when the distinction is not emphasized, the meanings expressed are consistent.

[0016] In the embodiments of the present application, sometimes the subscript such as W1 can be written in the form of non-subscript such as W1. When the distinction is not emphasized, the meanings expressed are consistent.

[0017] In order to make the technical problems, technical solutions and advantages of the present application clearer, the following will be described in detail with reference to the drawings and specific embodiments.

[0018] Reference is made to the drawings attached in the specification Figure 1 , which shows a flowchart of a temperature controller intelligent temperature regulating method provided by an embodiment of the present application.

[0019] This invention provides a method for intelligent temperature control of a temperature controller. This method can be implemented by an intelligent temperature control device, which can be a terminal or a server. The processing flow of the intelligent temperature control method may include the following steps: S1: Obtain the original multi-source temperature control dataset of the target controlled object.

[0020] The target controlled object refers to the specific temperature control system or spatial entity that the temperature controller acts on, the object used to implement temperature regulation and control, such as indoor rooms, building areas, equipment cabins, or industrial control spaces.

[0021] The original multi-source temperature control dataset refers to the basic data collection related to temperature regulation obtained from multiple different sources before preprocessing or fusion.

[0022] In one possible implementation, the original multi-source temperature control dataset includes: indoor temperature data (sensor temperature), outdoor temperature data, and historical control command data corresponding to the temperature controller.

[0023] Specifically, indoor temperature data refers to temperature data collected by a temperature sensor.

[0024] It should be noted that by acquiring the original multi-source temperature control dataset of the target controlled object, the actual operating state of the controlled space can be comprehensively and objectively characterized in the initial stage of the control system. This provides a sufficient information foundation for subsequent temperature reliability processing, predictive modeling, and control decisions. Compared with relying solely on data from a single temperature sensor, multi-source data can simultaneously reflect the coupling relationship between changes in the indoor thermal environment, external environmental disturbances, and the controller's own regulatory behavior, which is beneficial for revealing the intrinsic driving factors of temperature changes. Furthermore, preserving the original form of the data without excessive filtering or simplification in the early stages helps subsequent steps to perform targeted reliability assessments and weight allocations based on different needs, avoiding the loss of key information and thus improving the adaptability, robustness, and control accuracy of the overall temperature control model.

[0025] S2: Based on the neighborhood collaborative voting mechanism, the original multi-source temperature control dataset is made trustworthy to generate a trustworthy temperature field.

[0026] Among them, the neighborhood collaborative voting mechanism refers to a collaborative decision-making mechanism that takes temperature sensor nodes as the center, constructs a local neighborhood based on their spatial adjacency, and performs statistical analysis and consistency judgment on the temperature data of multiple sensors within the neighborhood, so that multiple interconnected nodes can jointly participate in the reliability assessment of temperature data.

[0027] Among them, the reliable temperature field refers to the spatial temperature distribution result composed of multiple temperature sensor nodes that have been determined to be reliable after the reliable processing is completed. It is used to truly reflect the overall temperature state of the controlled object at the current moment.

[0028] It should be noted that by introducing a trustworthiness processing method based on a neighborhood collaborative voting mechanism, the correlation of temperature sensors in spatial distribution can be fully utilized. This prevents individual sensor temperature data from being used in isolation for judgment; instead, multiple sensors within its neighborhood collaboratively evaluate its trustworthiness. This effectively reduces the impact of abnormal temperature data caused by sensor failure, local thermal disturbances, or transient noise on the overall temperature field modeling. Compared to traditional anomaly detection methods based on thresholds or simple statistical rules, this mechanism can dynamically adapt to the temperature distribution characteristics of different spatial regions, avoiding misjudging normal temperature data that exhibits gradient changes.

[0029] In one possible implementation, S2 specifically includes: S201: Based on the spatial location of the temperature sensor, the indoor temperature data is mapped into a temperature matrix according to a regular grid, where each element in the temperature matrix corresponds one-to-one with the corresponding sensor node.

[0030] The temperature matrix refers to a two-dimensional or multi-dimensional matrix structure formed by mapping discretely collected indoor temperature data onto a regular grid based on the actual spatial location of the temperature sensor in the controlled space. Each element in the matrix corresponds one-to-one with a specific temperature sensor node and is used to characterize the spatial temperature distribution.

[0031] S202: Using sensor nodes as the center, construct the Von Neumann neighborhood based on spatial index relationships.

[0032] Among them, the Von Neumann neighborhood refers to a local spatial neighborhood structure centered on a certain sensor node, which only includes its adjacent nodes in the upper, lower, left, and right of the regular grid, and is used to describe the local spatial relationship between sensors.

[0033] S203: Filter the Von Neumann neighborhood to determine the legitimate neighborhood alliance.

[0034] Among them, the legitimate neighborhood alliance refers to a cooperative set composed of a central node and its spatially adjacent sensor nodes, under the condition of satisfying the preset spatial index constraints.

[0035] Specifically, when constructing the neighborhood collaborative voting set, for any temperature sensor node in the temperature space matrix... Using this node as the neighborhood center, sequentially determine its relationship with other temperature sensor nodes. The relative positional relationship of nodes in matrix coordinate space. With nodes The coordinate index satisfies At that time, determine the node Located at node Within the Von Neumann neighborhood, and allows it to participate with nodes. A neighborhood-based cooperative alliance is formed. When the conditions are not met, the node is judged... Temperature sensor nodes that are not within the neighborhood are excluded from participating in the current collaborative alliance. This approach ensures that only temperature sensor nodes spatially adjacent to the central node are allowed to form legitimate neighborhood alliances, thus guaranteeing that the collaborative voting process is based on locally spatially consistent temperature data.

[0036] S204: Calculate the average temperature of each node within the legal neighborhood alliance.

[0037] S205: Calculate the root mean square deviation between the temperature of each node and the average value within the legal neighborhood alliance.

[0038] S206: Based on the root mean square deviation and the Student t-distribution, construct the temperature confidence interval: in, In the k-th cooperative iteration, s represents the sensor node s. i The constructed temperature confidence interval, Indicates sensor node s i The average temperature of the node and all its neighbors at the k-th iteration. Indicates the interval symmetric extension symbol. This represents the critical value of the Student t-distribution, i.e., at a degree of freedom of V⁻¹ and a significance level of 1. In the Student t-distribution, the critical value corresponding to the two-sided confidence interval, and V represents the number of sensor nodes participating in the collaboration within the neighborhood. Indicates sensor node s i It measures the degree of dispersion of all nodes in its neighborhood relative to the average temperature at the k-th iteration.

[0039] Among them, the temperature confidence interval refers to the reasonable temperature fluctuation range constructed based on the mean temperature, dispersion, and statistical characteristics of the Student t distribution within the neighborhood. It is used to measure the statistically reliable range of temperature data from a single sensor.

[0040] S207: Based on the temperature confidence interval, vote on the sensor nodes to determine the set of trustworthy nodes.

[0041] The set of trusted nodes refers to the set of sensor nodes that are deemed trustworthy during the collaborative voting process.

[0042] Specifically, during the collaborative voting process, for each temperature sensor node s i The currently collected temperature value and the corresponding temperature confidence interval Compare the temperature values. Falling within the temperature confidence interval Within this timeframe, the temperature sensor node is determined to meet the neighborhood statistical consistency condition, and its voting result is assigned accordingly. This indicates that the node is eligible to vote. When the temperature value... Not falling within the temperature confidence interval If the temperature sensor node does not meet the neighborhood statistical consistency condition, then assign it a voting result. This indicates that the node is not eligible to vote. By sequentially performing the above determination process on each temperature sensor node in the neighborhood, a voting vector composed of all voting results is formed. The temperature sensor nodes with a voting result of 1 are then aggregated into a set of trusted nodes for subsequent construction of the trusted temperature field.

[0043] S208: Repeat step S207 until the voting vectors of two adjacent iterations are consistent, and determine the reliable temperature field.

[0044] Specifically, when the following conditions are met A reliable temperature field is generated based on the final set of reliable nodes, where PV represents the reward vector. This represents the cooperative alliance in the k-th iteration. This represents the cooperative alliance in the (k+1)th iteration.

[0045] It should be noted that by fully considering the local consistency relationship between the node and its surrounding spatial nodes, and constructing temperature confidence intervals using root mean square deviation and Student t-distribution, the determination of temperature fluctuations has a clear statistical basis. This method can adapt to different neighborhood sizes and temperature dispersion levels, effectively distinguishing between normal fluctuations and abnormal deviations. By using iterative voting until the voting vector converges, the method avoids the randomness of one-time decisions, enhancing the stability and reliability of the trusted node selection results. Overall, this method can significantly suppress the interference of local abnormal sensors, instantaneous noise, or distorted data on temperature field modeling, providing a more stable, realistic, and spatially consistent temperature data foundation for subsequent temperature weight construction, predictive analysis, and control decisions.

[0046] S3: Calculate the temperature confidence weight matrix based on the confidence temperature field and the original temperature field.

[0047] The raw temperature field refers to the spatial distribution of indoor temperature data directly collected by all temperature sensors without any credibility processing, and includes possible noise, anomalies, or distorted data.

[0048] The temperature confidence weight matrix refers to the matrix structure formed by assigning corresponding confidence weights to each temperature sensor node based on the error relationship between the confidence temperature field and the original temperature field. It is used to quantify the confidence level and influence weight of different temperature data in the subsequent fusion, prediction and control process.

[0049] In one possible implementation, S3 specifically includes: S301: Calculate the temperature error matrix for each temperature node based on the reliable temperature field and the original temperature field.

[0050] The temperature error matrix refers to the error distribution matrix obtained by comparing the corresponding temperature nodes in the reliable temperature field and the original temperature field point by point. It is used to reflect the degree of deviation between the current temperature collected by each temperature sensor node and the reliable reference temperature.

[0051] S302: Based on the preset allowable error margin, correct the temperature error matrix and determine the correction error matrix.

[0052] The allowable error margin refers to the pre-set acceptable error range in temperature measurement and control applications, used to distinguish between normal measurement fluctuations and abnormal deviations that require special attention. Those skilled in the art can set the size of the allowable error margin according to actual needs; this invention does not limit this setting. The corrected error matrix refers to the error matrix obtained by correcting the original temperature error matrix with the allowable error margin. It weakens the influence of small normal errors, making the error evaluation more consistent with engineering practice.

[0053] Specifically, the formula for calculating the correction error matrix is ​​as follows: in, This represents the temperature error matrix at time t after correction with the allowable error margin, i.e., the correction error matrix. Let Id represent the temperature error matrix at time t, and let Id represent the identity matrix. This indicates the allowable error margin.

[0054] S303: Construct a centered scale based on the average error and maximum error of the correction error matrix.

[0055] The centralization scale refers to the scale parameter composed of the average error and the maximum error of the correction error matrix, which is used to describe the reference amplitude of the overall error distribution.

[0056] Specifically, the sum of the average error and the maximum error is the centering scale.

[0057] S304: Based on the centralized scale, the correction error matrix is ​​converted into a normalized percentage error matrix.

[0058] Specifically, in obtaining the correction error matrix and the corresponding centralized scale Then, for any element in the correction error matrix The normalized percentage error is calculated by normalizing the corrected error value with the centering scale and then multiplying it by a percentage factor of 100. The centralization scale is used to unify the dimensions and amplitude range of errors from different sensors, enabling the errors at each temperature node to be compared under the same reference standard. By performing the above calculation on all elements in the correction error matrix, a normalized percentage error matrix is ​​formed, which characterizes the relative deviation of each temperature node from the overall error level and serves as the basis for subsequent temperature reliability assessment and weight construction.

[0059] S305: Construct a reliable temperature weight matrix based on the normalized percentage error matrix.

[0060] It should be noted that the hierarchical modeling process described above, from temperature error calculation, error margin correction, centralized scale construction to error normalization, allows for an objective comparison of the error levels of different temperature sensor nodes under the same evaluation benchmark. This avoids imbalances in weight allocation caused by differences in sensor ranges or local extreme errors. Introducing an allowable error margin to correct the error matrix effectively filters out normal measurement fluctuations and highlights key errors that have a real impact on control performance, making the weight construction more aligned with engineering application scenarios. Centralized scales impose unified scale constraints on the errors and further transform them into normalized percentage error forms, ensuring good numerical stability and interpretability in the weight calculation process. The temperature reliability weight matrix constructed on this basis can continuously and precisely reflect the reliability differences of each temperature node, providing more reasonable and robust data weight support for subsequent temperature feature fusion, time-series prediction, and control decisions, thereby improving the accuracy and robustness of the overall temperature control system.

[0061] S4: Perform time-series alignment and feature fusion processing on the reliable temperature field and temperature reliable weight matrix to determine the input sequence.

[0062] Temporal alignment refers to unifying temperature and control data from different sources, with different sampling frequencies, or with time delays, to the same time reference through time index mapping or interpolation, ensuring consistency of multi-source data in the time dimension. Feature fusion processing, based on temporal alignment, involves weighted combination of reliable temperature field information and corresponding reliable weights, and integrating it with relevant control or environmental features to form a comprehensive feature representation that fully describes the system state. The input sequence refers to the multi-dimensional feature sequence arranged in chronological order after the above processing, used as input for subsequent time-series prediction and control models.

[0063] It should be noted that by performing unified temporal alignment and feature fusion processing on the reliable temperature field and the reliable temperature weight matrix, the inconsistencies in sampling time, update frequency, and response delay of multi-source temperature control data can be effectively eliminated. This ensures that various features participate in modeling on the same time scale. By constructing a structured input sequence, not only can the dynamic information of temperature evolution over time be completely preserved, but a clear and continuous data foundation is also provided for subsequent time-series prediction models to capture temperature change patterns and control response characteristics. This helps to improve the accuracy of temperature prediction and the stability and reliability of control decisions.

[0064] S5: Using a time-series prediction model based on thermodynamic constraints, the input sequence is processed through multi-step temperature prediction, and the temperature prediction sequence for the future control period is output.

[0065] Among them, the thermodynamically constrained time-series prediction model refers to a prediction model that, based on the traditional time-series prediction model structure, introduces indoor thermal balance relationships or physical constraints on temperature changes. This model is used to constrain temperature changes to meet basic thermodynamic laws during data-driven prediction. The temperature prediction sequence within the future control period refers to the sequence of indoor temperature prediction results output by the prediction model in chronological order within a preset control period, used to reflect the evolution trend of the controlled object's temperature over a future period.

[0066] In one possible implementation, S5 specifically includes: S501: Construct the input vector for the current time step based on the input sequence.

[0067] The input vector includes indoor temperature, historical control command mapping, and outdoor temperature.

[0068] S502: Based on the input vector and the indoor thermal balance relationship, establish the temperature change constraint equation: in, This represents the rate of change of indoor temperature over time. This represents the indoor equivalent heat capacity parameter. ( )express The power representation obtained by mapping historical control commands at any given time. Indicates the overall heat transfer coefficient. This represents the indoor temperature value collected by the temperature sensor at time t. This represents the outdoor ambient temperature at time t.

[0069] The indoor thermal balance relationship refers to a physical model describing the relationship between energy input, heat loss, and temperature change in an indoor space, used to characterize the fundamental thermodynamic laws governing temperature changes. The temperature change constraint equation is a differential equation established based on the indoor thermal balance relationship, used to constrain the physical rationality of indoor temperature changes over time.

[0070] S503: Input the input vector into the time-series prediction model based on thermodynamic constraints to determine the single-step indoor temperature prediction value.

[0071] Specifically, at the current time t, the indoor temperature value Power characterization quantity and outdoor ambient temperature The state input is fed into the time series prediction model, which calculates the temporal relationship between the input variables according to its internal mapping function f(), and obtains the next time step. Corresponding indoor temperature forecast value This allows for a single-step temperature prediction based on the current state, where... This indicates the prediction time step.

[0072] Among them, the single-step indoor temperature prediction value refers to the indoor temperature prediction result corresponding to the next time step calculated by the time series prediction model based on the current input vector.

[0073] Specifically, the time series prediction model is an LSTM model.

[0074] S504: Combine the temperature change constraint equation to perform consistency verification on the single-step indoor temperature prediction value.

[0075] Specifically, performing consistency verification on single-step predictions can ensure that the predicted trajectory satisfies the physical relationship of "power input - ambient temperature difference - room temperature change".

[0076] S505: Determine whether the consistency check result exceeds the threshold. If yes, correct the single-step prediction and proceed to step S506. Otherwise, proceed to step S507.

[0077] Those skilled in the art can set the threshold value according to actual needs, and the present invention does not limit it.

[0078] S506: Weighted correction is applied to the single-step indoor temperature prediction value based on the physical predicted temperature within the current control cycle. in, This represents the predicted temperature value after physical consistency correction. Indicates the weighting coefficient. This represents the physically predicted temperature calculated based on the temperature change constraint equation. This indicates the prediction time step.

[0079] S507: Using the single-step indoor temperature forecast as the initial value, and recursively extrapolating according to the forecast step size of the control cycle, the temperature forecast sequence for future control cycles is determined. in, This represents the temperature prediction sequence within the future control period. This represents the predicted indoor temperature at the k-th prediction time, where k = 1, 2, ..., N. p N p This indicates the prediction step size.

[0080] It should be noted that by introducing indoor thermal balance relationships and constructing temperature change constraint equations during the temperature prediction process, the prediction model is always constrained by physical laws when learning from historical data. This effectively avoids unreasonable predictions from purely data-driven models under complex operating conditions or sparse data. By performing physical consistency checks on single-step prediction results and introducing a weighted correction mechanism based on the physical model when deviations exceed thresholds, the stability and interpretability of prediction results can be improved while maintaining the flexibility of model prediction. The use of a rolling recursive approach to generate multi-step temperature prediction sequences ensures that the prediction results continuously reflect the temperature evolution trend within future control cycles. This provides a complete, continuous, and physically reasonable state prediction basis for subsequent predictive control models, facilitating the implementation of more forward-looking, stable, and energy-efficient temperature regulation and control strategies.

[0081] S6: Construct a fast analytical predictive control model based on the temperature prediction sequence.

[0082] Among them, the fast analytical predictive control model refers to a control model that determines the control input by optimizing the control objective, based on the prediction of the future state of the system and under the premise of satisfying the system constraints.

[0083] In one possible implementation, S6 specifically includes: S601: Use the temperature prediction sequence as the predicted state trajectory, and define the control variable as the control input increment vector: in, Indicates the current control moment Based on this, move forward to the first Predicted indoor temperature values ​​at each prediction step size Indicates time The incremental control input applied at point S i This represents the dynamic response parameter for the control input increment at step i. Indicates time The predicted temperature component is generated by the system's own dynamics and historical state evolution without considering the effect of the current control input increment.

[0084] in, The temperature prediction sequence within the future control period output in step S5 is used as a state component in the predicted state trajectory in step S601. Each predicted temperature state in the predicted state trajectory serves as a component of the predicted output vector in subsequent steps, and the control input increment vector serves as a decision variable for subsequent linear mapping relationships and objective function optimization.

[0085] Specifically, the predicted state trajectory refers to the state evolution path composed of the temperature prediction sequence within the future control cycle, used to describe the dynamic process of indoor temperature changing over time in the prediction time domain. The control input increment vector refers to the vector composed of the change in control input relative to the previous control moment, used to characterize the adjustment amplitude and direction of the temperature control actuator in the prediction time domain.

[0086] S602: Based on the system's dynamic characteristics, establish a linear mapping relationship between the predicted output vector and the control input increment vector: in, Let S represent the predicted output vector starting at time k+1, and let S represent the dynamic response matrix. This represents the control input increment vector at time k. This represents the uncontrolled prediction vector output by the system starting at time k+1.

[0087] Specifically, the system is a temperature regulation system composed of a controlled indoor space, a temperature control actuator, and environmental heat exchange factors. The system's dynamic characteristics are used to describe the dynamic response relationship between the control input increment and the indoor temperature output. For example, the system can be an indoor temperature dynamic system, a building thermodynamic system, or an HVAC temperature control system.

[0088] S603: Based on the linear mapping relationship, and combined with the system order and prediction step size, construct a dynamic matrix to describe the influence of the control increment on the predicted output.

[0089] The dynamic matrix refers to the parameter or matrix structure used to describe the degree of influence and dynamic characteristics of the control input increment on the indoor temperature output. It reflects the response relationship of the controlled object to the control action at different time steps.

[0090] S604: Based on the dynamic matrix and the deviation between the measured temperature and the predicted temperature at the current moment, the prediction output is compensated for the deviation. in, This represents the prediction output vector after bias compensation, located j prediction steps ahead from the current control time k. This represents the measured temperature at the current control time k. and model predicted temperature The deviation between them, where I represents the deviation propagation vector.

[0091] S605: Based on the compensation results, and under the constraints, a fast analytical predictive control model is constructed with the objective function of minimizing the overall temperature regulation performance index. in, Describe the objective function. represents the control input increment vector, and b represents the prediction bias vector. This represents the output error weighting matrix. R represents the uncertainty sensitivity suppression weight coefficient, and R represents the control increment weighting matrix. This represents the uncertainty sensitivity matrix.

[0092] Specifically, the comprehensive temperature regulation performance indicators include: the weighted tracking error of the indoor temperature in the prediction time domain relative to the reference temperature trajectory, the weighted change of the control input increment, and the sensitivity suppression index of the prediction output to model uncertainty.

[0093] For example, at the current control time k, based on the predicted temperature trajectory within the future control cycle obtained in step S604 after deviation compensation, this predicted temperature trajectory is compared with a preset reference temperature trajectory to construct a prediction error vector characterizing the temperature tracking performance. Simultaneously, constraints such as control input amplitude constraints, control input rate of change constraints, and upper and lower limits of indoor temperature are set according to the physical capabilities of the temperature control actuator and the system's operational safety requirements. Under these constraints, the predicted temperature deviation is weighted and accumulated within the prediction time domain, and penalty terms are introduced for the magnitude of the control input increment and the sensitivity of the predicted output to model uncertainty, respectively. An optimization function is constructed with the objective of minimizing the comprehensive temperature regulation performance index, which simultaneously reflects temperature tracking accuracy, control process stability, and system robustness to model uncertainty. Furthermore, a linear mapping relationship is established between the predicted output and the control input increment through the system dynamic response matrix, transforming the optimization problem into a predictive control model with a quadratic objective function and linear constraints, thereby enabling rapid analytical solution of the optimal control command within the current control cycle.

[0094] It should be noted that by explicitly constructing the future temperature prediction sequence as a predicted state trajectory and using the control input increment as the optimization variable, control decisions can be made proactively to plan for temperature change trends, rather than passively adjusting based solely on current errors. A linear mapping relationship between the predicted output and control input is established based on the system's dynamic characteristics, giving the control model a clear physical meaning and good interpretability. Introducing a deviation compensation mechanism effectively suppresses the accumulation and diffusion of model uncertainties and prediction errors in the prediction time domain, improving the stability of the control results. By constructing a comprehensive performance index that includes temperature tracking performance, control stability, and uncertainty sensitivity suppression terms, and transforming it into a predictive control model with analytical solutions, the optimal control command can be quickly solved under multiple constraints, thereby significantly improving the real-time performance, robustness, and overall control performance of the temperature control system in practical engineering applications.

[0095] In one possible implementation, the constraints include control input amplitude constraints, control input rate of change constraints, and indoor temperature upper and lower limit constraints.

[0096] S7: Solve the fast analytical predictive control model to determine the optimal control command within the current control cycle.

[0097] Among them, the optimal control command within the current control cycle refers to the temperature control command obtained by the predictive control model within the current control cycle, which enables the comprehensive temperature regulation performance index to reach the optimal or near-optimal level, and is used to drive the temperature control actuator to regulate the temperature of the controlled object.

[0098] In one possible implementation, S7 specifically includes: S701: Construct a prediction error vector based on the deviation between the reference temperature trajectory and the predicted uncontrolled output trajectory.

[0099] The reference temperature trajectory refers to the ideal indoor temperature change path predetermined in the prediction time domain based on the user-set target temperature or expected comfort requirements. The predicted uncontrolled output trajectory refers to the temperature prediction trajectory generated by the system's own dynamic characteristics and historical state evolution without applying current control input increments. The prediction error vector is a vector formed by the time-by-time differences between the reference temperature trajectory and the predicted uncontrolled output trajectory in the prediction time domain, used to quantify the degree to which the future temperature deviates from the target trajectory.

[0100] S702: Calculate the controller gain matrix by combining the dynamic matrix, the preset output error weighting matrix, and the control increment weighting matrix. Where, k c Let Q represent the controller gain matrix, Q represent the output error weighting matrix, and R represent the control increment weighting matrix. T Indicates transpose. -1 This represents the matrix inversion operation.

[0101] The output error weighting matrix is ​​a weighting matrix used to adjust the importance of temperature errors at different prediction times in the optimization objective. Those skilled in the art can set the size of the output error weighting matrix according to actual needs; this invention does not impose any limitations on this. The control increment weighting matrix is ​​a weighting matrix used to constrain the magnitude of control input changes and suppress over-adjustment. The controller gain matrix is ​​a gain matrix calculated jointly by the dynamic matrix, the output error weighting matrix, and the control increment weighting matrix, used to map the prediction error to the control input increment.

[0102] S703: Input the prediction error vector into the controller gain matrix to solve for the optimal control command within the current control cycle. in, This represents the optimal control input increment at the current control time k. This represents the prediction error vector in the prediction control time domain at the prediction start time k+1.

[0103] The optimal control input increment refers to the change in control input that, through optimization calculation, enables the overall control performance index to reach its optimal level within the current control cycle.

[0104] It should be noted that by constructing a prediction error vector based on the deviation between the reference temperature trajectory and the predicted uncontrolled output trajectory, control decisions can be directly corrected for future temperature deviation trends, rather than relying solely on the current time-instance error. This enhances the forward-looking nature of the control. Analytically calculating the controller gain matrix ensures a clear, stable, and repeatable mapping between the prediction error and the control input increment, avoiding the uncertainties introduced by complex iterative solutions. Directly mapping the prediction error vector to the optimal control input increment makes the control command generation process efficient and computationally manageable, which is beneficial for meeting the real-time and stability requirements of temperature controllers. Overall, this method can suppress drastic changes in control input while ensuring temperature tracking accuracy, thereby improving the operational stability, robustness, and reliability of the temperature control system in engineering applications.

[0105] S8: Execute the optimal control command and obtain the temperature feedback data after execution.

[0106] Temperature feedback data refers to the indoor temperature measurement data collected in real time by temperature sensors after the optimal control command is executed, which is used to reflect the actual temperature change of the controlled object under the control action.

[0107] It should be noted that by applying the optimal control command to the actual temperature control actuator and simultaneously acquiring the temperature feedback data after execution, a closed-loop connection between the control decision and the physical system can be achieved. This allows the control results to be directly reflected in real temperature changes. The temperature feedback data provides the system with an objective evaluation basis for the control effect, which is beneficial for timely detection of the impact of prediction errors, model deviations, or external disturbances. By continuously acquiring feedback temperature information, not only can the effectiveness of the control command within the current control cycle be verified, but it also provides real data support for subsequent weight updates, model corrections, and adaptive adjustments to the control strategy, thereby improving the stability, accuracy, and adaptability of the temperature control system during long-term operation.

[0108] In one possible implementation, the process after S8 includes: S9: Based on the temperature feedback data, update the temperature confidence weight matrix and the time series prediction model based on thermodynamic constraints, and return the update results to step S1.

[0109] Specifically, after acquiring the temperature feedback data following the execution of the optimal control command, the temperature feedback data is compared with the predicted temperature within the corresponding control cycle. The deviation between the actual and predicted temperatures is calculated. Based on the magnitude of the deviation and its distribution across different sensor nodes and time locations, the corresponding weights in the temperature reliability weight matrix are adaptively adjusted to reduce the impact of temperature data with large long-term deviations on subsequent fusion processes. Simultaneously, the operating state of the time-series prediction model is updated based on the temperature feedback data. By correcting the model's input state or adjusting its internal state parameters, the time-series prediction model can reflect the latest system operating characteristics, thereby continuously improving the accuracy of temperature prediction and the reliability of control decisions in subsequent control cycles.

[0110] In this embodiment of the invention, by introducing a neighborhood collaborative voting mechanism for the reliable processing of multi-source temperature data, a reliable temperature field is constructed and a reliable temperature weight matrix is ​​generated. This effectively suppresses the influence of abnormal sensor data and environmental noise on the temperature sensing results. Simultaneously, the reliable temperature information is fused with historical control data and outdoor environmental data over time, and thermodynamic constraints are introduced during temperature prediction to achieve high-precision, multi-step prediction of indoor temperature changes within future control cycles. Based on this, by constructing a rapid analytical predictive control model and solving for the optimal control command, the temperature regulation process, while meeting physical constraints and system operational safety requirements, achieves improved temperature tracking accuracy, a smoother control process, and a significant increase in system robustness and energy efficiency. This overall enhances the intelligent temperature regulation performance and engineering application reliability of the temperature controller under complex operating conditions.

[0111] Reference manual attached Figure 2 The diagram shows a structural schematic of an intelligent temperature control system for a temperature controller provided by the present invention.

[0112] The present invention also provides a temperature controller intelligent temperature control system 20, applied to the above-mentioned temperature controller intelligent temperature control method, comprising: Processor 201.

[0113] The memory 202 stores computer-readable instructions. When the computer-readable instructions are executed by the processor 201, they implement the intelligent temperature control method of the temperature controller as described in the method embodiment.

[0114] The intelligent temperature control system 20 provided by the present invention can execute the above-described intelligent temperature control method and achieve the same or similar technical effects. To avoid repetition, the present invention will not elaborate further.

[0115] It should be understood that the processor in the embodiments of the present invention can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0116] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0117] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0118] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0119] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0120] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply 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 invention.

[0121] Those skilled in the art will recognize that the units and algorithm steps of the various examples 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 implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0122] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0123] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0124] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0125] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0126] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0127] This invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the intelligent temperature control method for a temperature controller as described in the method embodiment.

[0128] The present invention provides a computer-readable storage medium that can implement the steps and effects of the intelligent temperature control method of the temperature controller in the above-described method embodiments. To avoid repetition, the present invention will not repeat the details.

[0129] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0130] The following points need to be explained: (1) The accompanying drawings of the embodiments of the present invention only involve the structures involved in the embodiments of the present invention. Other structures can refer to the general design.

[0131] (2) For clarity, the thickness of layers or regions is enlarged or reduced in the drawings used to describe embodiments of the invention, i.e., these drawings are not drawn to scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being “above” or “below” another element, the element may be “directly” located “above” or “below” the other element or there may be intermediate elements.

[0132] (3) Where there is no conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other to obtain new embodiments.

[0133] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for intelligent temperature control of a temperature controller, characterized in that, include: S1: Obtain the original multi-source temperature control dataset of the target controlled object; S2: Based on the neighborhood collaborative voting mechanism, the original multi-source temperature control dataset is processed to generate a reliable temperature field; S3: Calculate the temperature confidence weight matrix based on the confidence temperature field and the original temperature field; S4: Perform time-series alignment and feature fusion processing on the reliable temperature field and the reliable temperature weight matrix to determine the input sequence; S5: Using a time-series prediction model based on thermodynamic constraints, perform multi-step temperature prediction processing on the input sequence and output the temperature prediction sequence for the future control period. S6: Construct a fast analytical predictive control model based on the temperature prediction sequence; S7: Solve the fast analytical predictive control model to determine the optimal control command within the current control cycle; S8: Execute the optimal control command and obtain the temperature feedback data after execution.

2. The intelligent temperature control method for a temperature controller according to claim 1, characterized in that, The original multi-source temperature control dataset includes: indoor temperature data, outdoor temperature data, and historical control command data corresponding to the temperature controller.

3. The intelligent temperature control method for a temperature controller according to claim 1, characterized in that, S2 specifically includes: S201: Based on the spatial location of the temperature sensor, the indoor temperature data is mapped into a temperature matrix according to a regular grid, wherein each element in the temperature matrix corresponds one-to-one with the corresponding sensor node; S202: Using the sensor node as the center, construct the Von Neumann neighborhood based on spatial indexing relationships; S203: Filter the Von Neumann neighborhood to determine the legitimate neighborhood alliance; S204: Calculate the average temperature of each node within the legal neighborhood alliance; S205: Calculate the root mean square deviation between the temperature of each node in the legal neighborhood alliance and the average value; S206: Based on the root mean square deviation and the Student t distribution, construct the temperature confidence interval; S207: Based on the temperature confidence interval, vote on the sensor nodes to determine the set of trustworthy nodes; S208: Repeat step S207 until the voting vectors of two adjacent iterations are consistent, and determine the reliable temperature field.

4. The intelligent temperature control method for a temperature controller according to claim 1, characterized in that, S3 specifically includes: S301: Calculate the temperature error matrix for each temperature node based on the reliable temperature field and the original temperature field; S302: Based on the preset allowable error margin, the temperature error matrix is ​​corrected to determine the correction error matrix; S303: Construct a centered scale based on the average error and maximum error of the correction error matrix; S304: Based on the centralization scale, convert the correction error matrix into a normalized percentage error matrix; S305: Construct the temperature confidence weight matrix based on the normalized percentage error matrix.

5. The intelligent temperature control method for a temperature controller according to claim 1, characterized in that, S5 specifically includes: S501: Based on the input sequence, construct the input vector for the current moment; S502: Based on the input vector and the indoor thermal balance relationship, establish a temperature change constraint equation; S503: Input the input vector into the time-series prediction model based on thermodynamic constraints to determine the single-step indoor temperature prediction value; S504: Based on the temperature change constraint equation, perform a consistency check on the single-step indoor temperature prediction value; S505: Determine whether the consistency check result exceeds the threshold; if so, correct the single-step prediction and proceed to step S506; otherwise, proceed to step S507. S506: Combine the physical predicted temperature within the current control cycle to perform a weighted correction on the single-step indoor temperature prediction value; S507: Using the single-step indoor temperature prediction value as the initial value, the prediction step size of the control cycle is used for rolling recursion to determine the temperature prediction sequence within the future control cycle.

6. The intelligent temperature control method for a temperature controller according to claim 1, characterized in that, S6 specifically includes: S601: Use the temperature prediction sequence as the predicted state trajectory, and define the control variable as the control input increment vector; S602: Based on the dynamic characteristics of the system, establish a linear mapping relationship between the predicted output vector and the control input increment vector; S603: Based on the linear mapping relationship, and combined with the order of the system and the prediction step size, construct a dynamic matrix to describe the influence of the control increment on the prediction output; S604: Based on the dynamic matrix, and considering the deviation between the measured temperature and the predicted temperature at the current moment, the prediction output is compensated for the deviation. S605: Based on the compensation results, and under the constraints of the constraints, the fast analytical predictive control model is constructed with the objective function of minimizing the comprehensive temperature regulation performance index.

7. The intelligent temperature control method for a temperature controller according to claim 6, characterized in that, The constraints include control input amplitude constraints, control input rate of change constraints, and indoor temperature upper and lower limit constraints.

8. The intelligent temperature control method for a temperature controller according to claim 1, characterized in that, Specifically, S7 includes: S701: Construct a prediction error vector based on the deviation between the reference temperature trajectory and the predicted uncontrolled output trajectory; S702: Calculate the controller gain matrix by combining the dynamic matrix, the preset output error weighting matrix, and the control increment weighting matrix; S703: Input the prediction error vector into the controller gain matrix to solve for the optimal control command in the current control cycle.

9. The intelligent temperature control method for a temperature controller according to claim 1, characterized in that, Following S8, the following is also included: S9: Based on the temperature feedback data, update the temperature confidence weight matrix and the time series prediction model based on thermodynamic constraints, and return the update result to step S1.

10. A temperature controller intelligent temperature regulation system, characterized in that, include: processor; A memory storing computer-readable instructions, which, when executed by the processor, implement the intelligent temperature control method for a temperature controller as described in any one of claims 1 to 9.

Images