Multi-scene-oriented intelligent lamp lighting self-adaptive control method and system

Abstract

This invention discloses an adaptive control method and system for intelligent lighting fixtures in multiple scenarios, relating to the field of automatic control technology. The method includes acquiring the sampled sequence of controlled variables and the actuator feedback sequence for a controlled zone, generating a condition status fingerprint and a condition switching confidence value. Based on the condition status fingerprint and the condition switching confidence value, a deviation benchmark for the controlled variables in the zone is constructed, and cross-zone coupling control constraints are generated. An actuator control increment is generated based on the deviation benchmark and the cross-zone coupling control constraints, and the condition status fingerprint is updated using the actuator feedback sequence. This invention, through the closed-loop transmission of controlled variable sampling, actuator feedback, condition status fingerprint, deviation benchmark, cross-zone coupling constraints, control increment, and execution residual, incorporates the intelligent lighting fixture into a multi-zone automatic control process, realizing an adaptive, constrained, and feedback-updated control scheme for multiple scenarios.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology, specifically to an intelligent lighting adaptive control method and system for multiple scenarios. Background Technology

[0002] As the number of controlled objects in building spaces, public areas, and workplaces increases, automatic control systems for multi-zone objects are gradually shifting from single-loop control to multi-loop collaborative control. These systems typically acquire sampled values ​​of controlled variables for each controlled zone periodically through the controller, receive actuator control command readback values ​​and execution output feedback values, and then generate target control commands for the next control cycle based on the sampled state, execution state, and target trajectory. In intelligent lighting control objects, the controlled variables can manifest as lighting control quantities related to the zone's operating state. However, from the perspective of the control system, its core remains an automatic control process of sampling, judging, calculating, outputting, and updating feedback on controlled variables, actuator feedback, operating conditions, and control constraints.

[0003] Existing zonal control methods for multi-scenario control objects typically employ fixed target trajectories, fixed control step sizes, or feedback adjustment based on single sampling deviations. When multiple controlled zones participate in control simultaneously, changes in the controlled variable in a single region may be caused by a combination of real-world operating condition changes, external disturbances, actuator response lag, or coupling effects between adjacent zones. Existing methods struggle to accurately distinguish these state sources within the control cycle, easily misjudging short-term disturbances or execution delays as operating condition transitions. Furthermore, existing methods often directly use the difference between the target variable value and the current sampled value when generating control deviations, failing to adequately consider short-term sampling drift and target trajectory hold-up states before operating condition transition confirmation, leading to discontinuous changes in the deviation benchmark at boundary periods. For controlled zones with adjacency or coupling relationships, if each zone generates control increments independently, adjacent zones may exhibit opposite control directions, excessively large control increment differences, and continued adjustment despite insufficient actuator margin, resulting in control oscillations, accumulated response lag, and incorrect target trajectory hold-up. Therefore, there is a need for an adaptive control method that is oriented towards intelligent lighting control objects, but with multi-zone automatic control closed loop as the core, so as to coordinate the identification of operating conditions, construction of deviation benchmarks, generation of cross-zone constraints, allocation of actuator control increments and feedback status updates within the control cycle. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is that existing multi-scene adaptive lighting control methods have the following problems: it is difficult to distinguish between real working condition switching and external disturbances; the zone control deviation is prone to sudden changes due to short-term sampling drift; the independent adjustment of adjacent controlled zones leads to mutual interference of control increments; and how to form an iteratively updated multi-zone closed-loop control with the participation of actuator feedback.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: an intelligent lighting adaptive control method for multiple scenarios, comprising acquiring the controlled variable sampling sequence and actuator feedback sequence of the controlled zone, and generating a working condition fingerprint and a working condition switching confidence value. Based on the operating condition fingerprint and the operating condition switching confidence value, a deviation benchmark for the controlled variable in each partition is constructed, and cross-partition coupled control constraints are generated. The actuator control increment is generated based on the deviation benchmark of the controlled variable in the partition and the cross-partition coupled control constraint, and the operating condition fingerprint is updated using the actuator feedback sequence.

[0007] As a preferred embodiment of the intelligent lighting adaptive control method for multiple scenarios described in this invention, the operating condition fingerprint includes: receiving sampled values ​​of controlled variables according to the controlled zone identifier and forming a controlled variable sampling sequence according to the sampling time sequence; receiving actuator control command readback values ​​and execution output feedback values ​​according to the actuator identifier and forming an actuator feedback sequence according to the control cycle; performing time-series smoothing on the controlled variable sampling sequence and determining the direction of change of the controlled variables; determining the actuator response hysteresis flag according to the actuator control command readback values ​​and execution output feedback values; and writing the controlled zone identifier, the direction of change of the controlled variables, and the actuator response hysteresis flag into the operating condition fingerprint.

[0008] As a preferred embodiment of the intelligent lighting adaptive control method for multiple scenarios described in this invention, the working condition switching confidence quantity includes: comparing the direction of change of the controlled variable, the actuator response lag mark, and the interval change synchronization relationship in the fingerprints of adjacent working conditions within a continuous control cycle. The interval change synchronization relationship represents the degree of consistency between the trend of change of the controlled variable and the actuator response state between adjacent controlled zones. When the direction of change of the controlled variable is inconsistent with the actuator response lag mark, the corresponding control cycle is marked as an external disturbance cycle. Working condition switching evidence is accumulated in the control cycle that has not been marked as an external disturbance cycle. When the working condition switching evidence meets the switching confirmation condition and continues to meet the preset holding period, a working condition switching confidence quantity is generated. When the working condition switching evidence does not meet the switching confirmation condition, the working condition switching confidence quantity of the previous control cycle is maintained.

[0009] As a preferred embodiment of the intelligent lighting adaptive control method for multiple scenarios described in this invention, the step of constructing the partitioned controlled variable deviation benchmark includes: matching the trajectory of the target variable based on the operating condition fingerprint; generating the partitioned target variable value according to the trajectory of the target variable when the operating condition switching confidence value meets the switching confirmation condition; using the partitioned target variable value from the previous control cycle when the operating condition switching confidence value does not meet the switching confirmation condition; generating a time window compensation amount based on the change of the controlled variable sampling sequence within the sliding time window; superimposing the partitioned target variable value and the time window compensation amount; and subtracting the difference between the current sampled value of the controlled variable sampling sequence and the current sampled value to obtain the partitioned controlled variable deviation benchmark. The partitioned controlled variable deviation benchmark represents the control deviation between the target variable value and the actual sampled value within the controlled partition after time window compensation.

[0010] As a preferred embodiment of the intelligent lighting adaptive control method for multiple scenarios described in this invention, the generation of cross-zone coupling control constraints includes: establishing partition coupling edges based on the adjacency relationship or coupling influence relationship of the controlled partitions; the partition coupling edges characterize the mutual influence relationship between adjacent controlled partitions on the changes of controlled variables or actuator outputs; for each partition coupling edge, reading the partition controlled variable deviation benchmark of the two controlled partitions and calculating the deviation benchmark difference between the two controlled partitions; determining the actuator response margin of the two controlled partitions based on the actuator control command readback value and the execution output feedback value; determining the control direction consistency of the two controlled partitions based on the deviation benchmark difference; when the control direction consistency indicates that the control directions of the two controlled partitions are opposite, using the actuator response margin to limit the control increment difference between the two controlled partitions; and writing the limiting result of the control increment difference into the cross-zone coupling control constraint.

[0011] As a preferred embodiment of the intelligent lighting adaptive control method for multiple scenarios described in this invention, the generation of actuator control increments includes: converting the deviation benchmark of the controlled variable in the partition into a candidate control increment for the partition; performing consistency trimming on the candidate control increments of adjacent controlled partitions according to the cross-partition coupling control constraint; ensuring that the difference in control increments between adjacent controlled partitions meets the control boundary defined by the cross-partition coupling control constraint; using the trimmed candidate control increments and actuator response margin as inputs; allocating actuator control increments to the actuators in the corresponding controlled partitions; superimposing the actuator control increments onto the current control command of the actuator to form a target control command; outputting the target control command to the corresponding actuator; and writing the target control command into the actuator feedback sequence of the next control cycle.

[0012] As a preferred embodiment of the intelligent lighting adaptive control method for multiple scenarios described in this invention, the updated operating condition fingerprint includes: after the target control command is output, receiving a new execution output feedback value and a new controlled variable sample value; calculating the execution residual based on the target control command, the new execution output feedback value, and the new controlled variable sample value; the execution residual is used to characterize the degree of deviation between the expected response corresponding to the target control command and the execution output feedback and the actual change of the controlled variable; writing the execution residual back to the actuator response hysteresis marker and the interval change synchronization relationship in the operating condition fingerprint; when the execution residual is inconsistent with the switching direction corresponding to the operating condition switching confidence value, reducing the operating condition switching evidence; when the execution residual is consistent with the switching direction corresponding to the operating condition switching confidence value, retaining the operating condition target variable trajectory.

[0013] As a preferred embodiment of the intelligent lighting adaptive control system for multiple scenarios described in this invention, it includes: a fingerprint switching module, a deviation coupling module, and an incremental write-back module. The fingerprint switching module is used to acquire the controlled variable sampling sequence and actuator feedback sequence of the controlled partition, and generate the operating condition fingerprint and operating condition switching confidence value. The deviation coupling module is used to construct the deviation benchmark of the controlled variable in each partition based on the operating condition fingerprint and the operating condition switching confidence value, and to generate cross-partition coupled control constraints. The incremental write-back module generates actuator control increments based on the deviation benchmark of the controlled variable in the partition and the cross-partition coupled control constraints, and updates the operating condition fingerprint using the actuator feedback sequence.

[0014] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement a method for adaptive control of intelligent lighting fixtures for multiple scenarios.

[0015] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of an intelligent lighting adaptive control method for multiple scenarios.

[0016] The beneficial effects of this invention are: By forming a controlled variable sampling sequence according to the controlled zone identifier and an actuator feedback sequence according to the actuator identifier, and writing the controlled variable change direction, actuator response lag mark and interval change synchronization relationship into the operating condition fingerprint, the synchronous representation of the controlled zone operating status and actuator response status is realized. This enables the operating condition switching judgment to exclude external disturbance cycles and accumulate effective switching evidence, thereby achieving the beneficial effects of reducing erroneous operating condition switching and improving the reliability of control state input.

[0017] By matching the trajectory of the target variable based on the operating condition fingerprint, and generating a time window compensation amount using a sliding time window when the confidence value of the operating condition switching does not meet the switching confirmation condition, the continuous connection between the target variable value of the partition and the short-term sampling change is realized. This allows the deviation benchmark of the controlled variable of the partition to reflect the target tracking requirements and short-term change trends. At the same time, by limiting the control increment difference between adjacent partitions through cross-partition coupling control constraints, the beneficial effect of reducing mutual interference and boundary oscillation between adjacent control loops is achieved.

[0018] By converting the deviation benchmark of the controlled variable in a partition into a candidate control increment for that partition, and performing consistent pruning based on the cross-partition coupling control constraints, the coordination between the control increment and the coupling boundary of adjacent partitions is achieved. This ensures that the target control command satisfies both the actuator response margin and the cross-partition control boundary. Furthermore, by executing residual write-back of the operating condition fingerprint, the beneficial effect of continuously correcting the operating condition judgment and maintaining the target trajectory state using feedback results is achieved. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is an overall flowchart of the intelligent lighting adaptive control method for multiple scenarios provided in Embodiment 1 of the present invention. Detailed Implementation

[0021] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail 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 them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0022] Example 1, referring to Figure 1 As one embodiment of the present invention, an adaptive control method for intelligent lighting fixtures oriented towards multiple scenarios is provided, including: S1: Obtain the controlled variable sampling sequence and actuator feedback sequence of the controlled partition, and generate the operating condition fingerprint and operating condition switching confidence value.

[0023] The operating condition fingerprint includes: receiving the sampled values ​​of the controlled variable according to the controlled zone identifier and forming a sampled sequence of the controlled variable according to the sampling timing; receiving the actuator control command readback value and execution output feedback value according to the actuator identifier and forming an actuator feedback sequence according to the control cycle; performing time-series smoothing on the sampled sequence of the controlled variable and determining the direction of change of the controlled variable; determining the actuator response hysteresis mark based on the actuator control command readback value and execution output feedback value; and writing the controlled zone identifier, the direction of change of the controlled variable, and the actuator response hysteresis mark into the operating condition fingerprint.

[0024] Furthermore, this embodiment is applied to an intelligent lighting adaptive control system. The controlled area is a space control unit that is independently sampled and issued control commands by the controller, such as a meeting area, passageway area, display area, work area, or public area. The sampled values ​​of the controlled variables are control quantity sampled values ​​that reflect the lighting operation status of the controlled area, including illuminance sampled values, luminance evaluation values, color temperature deviation, or lighting target deviation formed after sensor sampling. All of the above variables participate in closed-loop control as controlled variables in the control system.

[0025] Furthermore, the controlled zone identifier is used to distinguish different control loops in the control system, denoted as the zone number. The sampling sequence is the order in which the controller receives the sampled values ​​of the controlled variable according to a fixed sampling period. In this embodiment, the sampling period is set to 0.5 seconds to 5 seconds, preferably 1 second. This range is derived from the engineering preset boundary between the sensor response time and the controller calculation cycle in the indoor lighting control system. This setting ensures that changes in the controlled variable are captured in a timely manner, while avoiding excessive sampling that would increase the computational load on the controller. Each controlled zone generates at least one sampled value of the controlled variable within the same control cycle, and the sampled values ​​from at least 5 consecutive control cycles form a controlled variable sampling sequence.

[0026] Furthermore, the actuator is an execution unit that receives target control commands from the controller and adjusts the operating state of the controlled zone lighting. The actuator feedback sequence includes the actuator control command readback value and the execution output feedback value. The actuator control command readback value represents the actual received command value returned by the actuator or actuator interface after the controller issues the command, and the execution output feedback value represents the output status feedback value after the actuator executes the command. Both are written into the feedback sequence with the control cycle as the time index, and their values ​​are uniformly normalized to the range of 0 to 1, where 0 indicates that the actuator is at the lowest control output boundary and 1 indicates that the actuator is at the highest control output boundary. The normalization upper and lower limits are derived from the engineering preset values ​​of the actuator's allowable control range. This setting allows the feedback quantities of different types of actuators to be compared in the same control logic.

[0027] Furthermore, temporal smoothing is used to suppress misjudgments caused by short-term sensor noise. Considering that factors such as personnel movement, obstruction, and sudden changes in natural light can cause short-term fluctuations in the sampled values ​​of the controlled variable in lighting control scenarios, this embodiment uses exponential smoothing to generate smoothed sampled values ​​and determines the direction of change of the controlled variable based on the difference between adjacent smoothed sampled values. in, Indicates the first The controlled partition in the first The smoothed sampled value for each control cycle is derived from the current sampled value and the smoothed sampled value of the previous control cycle, and its range is consistent with the sampled value of the controlled variable. Indicates the first The controlled partition in the first The sampled values ​​of the controlled variable for each control cycle are derived from the controlled partition sensor or the sampling interface of the control system, and the values ​​are normalized to be 0 to 1. Indicates the first The smoothing coefficient of each controlled partition is set to 0.25 to 0.45 in this embodiment, which is derived from the 75th percentile of the sampling fluctuation amplitude of adjacent control cycles in the calibration sample. This setting takes into account both the change response speed and noise suppression capability. Indicates the first The controlled partition in the first The direction of change of the controlled variable in each control cycle takes the value 1, 0 or -1, which represents rising, stable or falling, respectively. Indicates the first The boundary for determining the direction of change of each controlled zone is set to 0.01 to 0.05 in this embodiment. This value is derived from the larger value of the sampling resolution of the controlled variable and the static fluctuation range of the calibration sample. This setting avoids misjudging sampling jitter as a change in the actual control state. Indicates the controlled partition number, with a value ranging from 1 to... ,in The total number of controlled zones registered in the control system. This represents the control cycle number, and its value is a positive integer. In the formula result, The operating condition fingerprint is directly written to the system for subsequent comparison of the direction of change of the controlled variable in adjacent operating condition fingerprints.

[0028] Furthermore, the actuator response lag flag is used to characterize whether the actual output feedback of the actuator fails to keep up with the control command readback value in a timely manner after receiving the control command. Specifically, the difference between the actuator control command readback value and the execution output feedback value is taken as the command execution deviation. When this deviation exceeds the response deviation threshold for 2 to 4 consecutive control cycles, the corresponding actuator is recorded as having response lag. In this embodiment, the response deviation threshold is set to 0.03 to 0.08, derived from the engineering preset range of actuator feedback resolution and control interface readback error. This setting can distinguish between true actuator lag and single-cycle communication jitter. When there are multiple actuators in the same controlled partition, a partition-level actuator response lag flag is generated based on the proportion of the number of actuators with response lag in that partition to the total number of actuators. When the proportion is greater than 0.5, it is recorded as 1; otherwise, it is recorded as 0.

[0029] Within a continuous control cycle, the direction of change of controlled variables, actuator response lag marker, and interval change synchronization relationship in the fingerprints of adjacent operating conditions are compared. The interval change synchronization relationship is used to characterize the degree of consistency between the trend of change of controlled variables and the actuator response state between adjacent controlled zones. When the direction of change of controlled variables is inconsistent with the actuator response lag marker, the corresponding control cycle is marked as an external disturbance cycle. Operating condition switching evidence is accumulated in control cycles that have not been marked as external disturbance cycles. When the operating condition switching evidence meets the switching confirmation condition and continues to meet the preset holding period, an operating condition switching confidence value is generated. When the operating condition switching evidence does not meet the switching confirmation condition, the operating condition switching confidence value of the previous control cycle is maintained.

[0030] Furthermore, the interval change synchronization relationship applies to adjacent controlled zones to determine whether the changing trends of controlled variables between different control loops are in the same direction, similar, or opposite. The adjacency relationship of controlled zones can be given by the partition configuration table during control system initialization. The partition configuration table includes the partition number, adjacent partition numbers, and the correspondence of control loops. This configuration table belongs to the engineering preset data. For each pair of adjacent controlled zones, if the changing directions of their controlled variables are the same and the actuator response lag markers are the same, the synchronization relationship takes a higher value. If the changing directions are opposite or one zone shows a response lag while the other does not, the synchronization relationship takes a lower value. The synchronization relationship value ranges from 0 to 1, where 0 indicates asynchrony and 1 indicates complete synchronization.

[0031] Furthermore, the external disturbance period is used to exclude sampling changes not caused by the control system's own control actions. For example, in lighting control scenarios, factors such as people blocking the light, changes in ambient light caused by opening doors and windows, or a sudden occupancy of an adjacent area can all cause a mismatch between the direction of change of the controlled variable and the actuator response lag flag. Specifically, when the direction of change of the controlled variable indicates a rapid change, but the actuator feedback sequence does not show a corresponding change in control output, or when the actuator feedback shows a significant change but the controlled variable does not show a corresponding response, this control period is marked as an external disturbance period. The external disturbance period does not participate in the accumulation of operating condition switching evidence, but it is still retained in the sampling sequence as a reference for subsequent feedback updates.

[0032] Furthermore, the operating condition switching evidence is the amount of state change evidence accumulated in the control cycle not marked as an external disturbance cycle, and its object is the control association area formed by the current controlled partition and its adjacent controlled partitions. Considering that single-cycle changes are easily affected by sampling noise and control lag interference, this embodiment uses consistency of change direction, response lag stability, and inter-regional change synchronization relationship to jointly form the operating condition switching evidence: in, Indicates the first The controlled partition in the first The evidence for the switching of operating conditions in each control cycle comes from the comparison results of the operating condition fingerprints of the current cycle and the previous cycle, with values ​​ranging from 0 to 1. This function represents a truncation function that restricts the result within the parentheses to the range of 0 to 1, taking 0 if the result is less than 0 and 1 if the result is greater than 1. The evidence retention coefficient is set to 0.55 to 0.75 in this embodiment. It is derived from the engineering preset boundary of the minimum duration of operating condition changes in the control system. This setting avoids direct triggering of operating condition switching by single-cycle jumps. Indicates the first Evidence for the switching of operating conditions of each controlled zone in the previous control cycle comes from the calculation results of the previous control cycle. Indicates the first The controlled partition in the first The external disturbance cycle flag for each control cycle, with a value of 0 or 1. 1 indicates that the cycle is flagged as an external disturbance cycle, and 0 indicates that it is not flagged. The weight representing the consistency of the direction of change of the controlled variable is set to 0.3 to 0.45 in this embodiment, which is derived from the preset value of the contribution ratio of the control variable to the switching of operating conditions. The value represents the stability weight of the actuator response hysteresis. In this embodiment, it is set to 0.2 to 0.35, which is derived from the preset value of the contribution ratio of the execution feedback to the control state judgment. The weight of the synchronization relationship between the interval changes is set to 0.25 to 0.4 in this embodiment, which is derived from the engineering preset ratio of the coupling effect of adjacent control loops. This represents the sum of three weights, and the rule is set to equal 1, ensuring that the evidence for switching operating conditions remains dimensionless and consistent. Indicates the first The controlled partition in the first The direction of change of the controlled variable in each control cycle is derived from the aforementioned time-series smoothing calculation results. Indicates the first The direction of change of the controlled variable in each controlled partition in the previous control period. Indicates the first The controlled partition in the first The actuator response hysteresis flag for each control cycle is derived from the deviation judgment between the actuator control command readback value and the execution output feedback value, and takes a value of 0 or 1. Indicates the first The control partition is marked with a lag in the actuator response of the previous control cycle. Indicates the first The controlled partition and its adjacent controlled partitions in the The synchronization relationship of the interval changes in each control cycle is derived from the comparison results of the change direction and response lag mark of adjacent partitions, with a value range of 0 to 1.

[0033] In the formula results, Used for comparison with switching confirmation conditions and as input for generating operating condition switching confidence values.

[0034] Furthermore, the switching confirmation condition is that the evidence for the change in operating condition reaches or exceeds the evidence threshold and is continuously satisfied for a preset holding period. In this embodiment, the evidence threshold is set to 0.65 to 0.8, derived from the intermediate dividing boundary between the distribution of stable operating condition and switching operating condition evidence in the calibration sample. The preset holding period is set to 3 to 8 control cycles, derived from the minimum identifiable duration of the lighting control system under typical operating conditions such as personnel entry, exit, and changes in ambient light. This setting prevents false triggering of boundary samples and ensures that the true operating condition switch can be confirmed within an acceptable control delay. The generated operating condition switching confidence value is between 0 and 1. When the operating condition switching evidence meets the switching confirmation condition, the confidence value increases with the number of consecutively satisfied cycles. When the condition is not met, the operating condition switching confidence value of the previous control cycle is maintained.

[0035] Furthermore, the operating condition switching confidence level is not a standalone identification conclusion, but rather a state input quantity within the control closed loop. Its output serves two purposes: firstly, it is used in S2 for matching the trajectory of the operating condition target variable and generating the values ​​of the partition target variable; secondly, it is compared with the execution residual in S3 to determine whether to reduce the operating condition switching evidence or retain the trajectory of the operating condition target variable. If there is a sampling gap in the current control cycle, the direction of change of the controlled variable from the previous control cycle is followed, and the operating condition switching evidence for this cycle remains unchanged. If the synchronization relationship between adjacent partitions cannot be obtained, the synchronization relationship is set to 0.5, indicating that the switching evidence is neither strengthened nor weakened, thus ensuring that the control flow can still be executed even with missing inputs.

[0036] Furthermore, in this embodiment, the data structure of the operating condition fingerprint includes at least the controlled partition identifier, the direction of change of the controlled variable, the actuator response lag marker, the synchronization relationship of changes between partitions, the external disturbance cycle marker, and the operating condition switching evidence. All of the above data originates from the controlled variable sampling sequence, the actuator feedback sequence, the controlled partition adjacency configuration table, or the output result of the previous control cycle, without introducing any unavailable external judgment objects. This operating condition fingerprint is updated with each control cycle, serving as the preceding state input for S2 to construct the partition controlled variable deviation benchmark and the cross-partition coupled control constraints.

[0037] It should be noted that this step is not just a simple filtering of the lighting sample values, but rather the control variable change direction, actuator response lag marker, external disturbance period exclusion, and interval change synchronization relationship are all written into the operating condition fingerprint, so that subsequent control judgments can simultaneously consider the state of the controlled object, the actuator response state, and the synchronization state of adjacent control loops. This processing method solves the problem that single sample values ​​are easily affected by disturbances in multi-scene lighting control, and that single execution feedback is difficult to judge the switching of operating conditions, and has obvious closed-loop control attributes.

[0038] S2: Construct the partitioned controlled variable deviation benchmark based on the operating condition fingerprint and operating condition switching confidence, and generate cross-partition coupled control constraints.

[0039] The process of constructing the partitioned controlled variable deviation benchmark includes: matching the trajectory of the target variable based on the operating condition fingerprint; generating the partitioned target variable value according to the trajectory of the target variable when the operating condition switching confidence level meets the switching confirmation condition; and using the partitioned target variable value from the previous control cycle when the operating condition switching confidence level does not meet the switching confirmation condition. A time window compensation amount is generated based on the change in the controlled variable sampling sequence within the sliding time window. The partitioned target variable value and the time window compensation amount are then superimposed, and the difference is calculated with the current sampled value of the controlled variable sampling sequence to obtain the partitioned controlled variable deviation benchmark. The partitioned controlled variable deviation benchmark is used to represent the control deviation between the target variable value and the actual sampled value within the controlled partition after time window compensation.

[0040] Furthermore, the target variable trajectory is a preset target control curve of the control system for different lighting usage scenarios, and its object is the controlled variable within the controlled zone. The target variable trajectory can be generated by the initial configuration of the control system, including the scene number, the initial value of the target variable, the stable value of the target variable, the transition time, and the allowable deviation range. Its source is the control strategy table preset by the project, and it does not involve the lamp circuit or light source structure. The values ​​of the target variable trajectory are uniformly normalized to the interval between 0 and 1, where 0 indicates that the controlled variable is at the lowest target boundary, and 1 indicates that it is at the highest target boundary. This setting allows different lighting usage scenarios such as meetings, corridors, displays, and work to be expressed as target variable trajectories in the same control system.

[0041] Furthermore, when matching the target variable trajectory based on the operating condition fingerprint, the controller reads the controlled zone identifier, the direction of change of the controlled variable, the actuator response hysteresis flag, and the operating condition switching confidence value, and selects the target trajectory closest to the current operating condition from the preset target trajectory table. The matching rules are as follows: when the operating condition switching confidence value meets the switching confirmation condition, the target variable trajectory corresponding to the switching direction is used; when the actuator response hysteresis flag is 1, the target variable trajectory with a smaller transition slope is preferred; when the direction of change of the controlled variable is in a stable state, the hold-type target variable trajectory is preferred. The target trajectory table includes at least the trajectory number, applicable zone type, target variable change direction, and target variable stable value, all of which are preset engineering data that the controller can read.

[0042] Furthermore, the partition target variable value is the target control value corresponding to the trajectory of the operating condition target variable in the current control cycle. Considering that the operating condition switching confidence value may fluctuate repeatedly during boundary periods, this embodiment does not directly abruptly change the target variable to the trajectory stable value. Instead, it gradually generates the partition target variable value according to the trajectory transition duration. The transition duration is set to 5 to 60 control cycles, derived from the comfort control boundary of the lighting control object to changes in brightness or illuminance and the response capability of the control system. This setting avoids excessive control increments caused by abrupt changes in the target value. If the operating condition switching confidence value does not meet the switching confirmation condition, the partition target variable value from the previous control cycle is used to ensure the continuity of the control target.

[0043] Furthermore, a sliding time window is used to calculate the change in the sampled sequence of the controlled variable within recent control periods. The sliding time window length is set to 5 to 20 control periods, and the sliding step size is set to 1 control period. This is derived from the minimum number of samples required for the control system to identify short-term trends and the boundary of real-time control computation. This setting can identify short-term drift of the controlled variable without masking current operating condition changes due to an excessively long time window. The time window compensation amount is applied to the partition target variable value to provide gentle compensation for short-term sampling drift when the confidence level for operating condition switching is not confirmed.

[0044] Furthermore, considering that the target variable value, current sampled value, and time window change amount need to form a reproducible control deviation input, this embodiment generates the time window compensation amount and the partitioned controlled variable deviation benchmark according to the following formula: in, Indicates the first The controlled partition in the first The time window compensation amount for each control cycle is derived from the change in the sampled sequence of the controlled variable within the sliding time window, and its value is truncated to -0.2 to 0.2. Indicates the first The time window compensation coefficient for each controlled partition is set to 0.3 to 0.6 in this embodiment. It is derived from the engineering preset ratio of the impact of short-term sampling drift on control deviation in the calibration sample. This setting prevents the time window compensation amount from being too large and replacing the target control value itself. Indicates the first The controlled partition in the first The current sampled value for each control cycle is derived from the sampled sequence of the controlled variable, and the normalized value ranges from 0 to 1. Indicates the first The sampled values ​​of the controlled variables at the starting point of the sliding time window for each controlled partition originate from the same sampled sequence of controlled variables. Indicates the first The sliding time window length for each controlled partition is set to 5 to 20 control periods in this embodiment, and not less than 2 when the number of available samples is insufficient. When calculating, use the existing sample size and ensure that the denominator is not less than 1. Indicates the first The controlled partition in the first The deviation benchmark of the controlled variable in each control cycle is derived from the target variable value of the zone, the time window compensation amount, and the current sampled value, and the value is truncated to -1 to 1. Indicates the first The controlled partition in the first The target variable values ​​for each control cycle are derived from the trajectory of the target variable under operating conditions or the target variable values ​​for the previous control cycle, and are normalized to values ​​between 0 and 1. In the formula results, As the basic input for generating subsequent cross-regional coupling control constraints and actuator control increments, When positive, it indicates that the target variable value is higher than the actual sampled value, and control needs to be increased in the direction of the controlled variable. When the value is negative, it indicates that the target variable value is lower than the actual sampled value, and control needs to be applied in the direction of reducing the controlled variable.

[0045] Based on the adjacency or coupling influence relationship of the controlled partitions, partition coupling edges are established. These edges characterize the mutual influence relationship between changes in controlled variables or actuator outputs between adjacent controlled partitions. For each partition coupling edge, the deviation benchmarks of the controlled variables at both ends of the controlled partitions are read, and the difference between the deviation benchmarks at both ends is calculated. The actuator response margin at both ends of the controlled partitions is determined based on the actuator control command readback value and the execution output feedback value. The control direction consistency between the two ends of the controlled partitions is determined based on the deviation benchmark difference. When the control direction consistency indicates that the control directions of the two ends of the controlled partitions are opposite, the control increment difference between the two ends of the controlled partitions is limited using the actuator response margin. The limitation result of the control increment difference is written into the cross-partition coupling control constraint quantity.

[0046] Furthermore, a partition coupling edge is a control association object between two controlled partitions in the control system, established based on the adjacency relationship or coupling influence relationship of the controlled partitions. The adjacency relationship of the controlled partitions comes from the partition configuration table, such as two controlled partitions having spatially adjacent boundaries, sharing channels, or adjacent control ranges within the same open space. The coupling influence relationship comes from the calibration configuration during the initialization of the control system, such as a change in the output of an actuator in one controlled partition having a measurable impact on the sampled values ​​of the controlled variables in adjacent controlled partitions. Each partition coupling edge records the partition numbers at both ends, the coupling weight, and the allowable control increment difference boundary.

[0047] Furthermore, the deviation benchmark difference is the difference between the deviation benchmarks of the controlled variables of the controlled partitions at both ends of the same partition coupling edge. It is used to determine whether there are significant differences in the control requirements of adjacent controlled partitions. The consistency of control direction is determined based on the signs of the deviation benchmarks of the controlled variables at both ends of the partitions: when both deviation benchmarks are positive or both are negative, it indicates that the control direction is consistent; when one end is positive and the other end is negative, it indicates that the control direction is opposite; when the absolute value of the deviation benchmark at either end is less than the direction dead zone boundary, it indicates that the control direction at that end is uncertain. In this embodiment, the direction dead zone boundary is set to 0.02 to 0.06, derived from the sampling accuracy of the controlled variable and the allowable deviation range of the control target. This setting avoids frequent reversals of the control direction between adjacent partitions due to small deviations.

[0048] Furthermore, the actuator response margin characterizes the available space for the actuator to further increase or decrease the control output within the current control cycle. For control directions requiring an increase in the controlled variable, the actuator response margin is determined by the difference between the highest control boundary and the current execution output feedback value of the actuator. For control directions requiring a decrease in the controlled variable, the actuator response margin is determined by the difference between the current execution output feedback value of the actuator and the lowest control boundary. The highest and lowest control boundaries are set to normalized 1 and 0, respectively, derived from the actuator's permissible control range. This setting ensures that the controller will not exceed the actual executable range of the actuator when generating adjacent zone control increment limits.

[0049] Furthermore, considering that adjacent partitions with opposite control directions are prone to mutual cancellation, repeated adjustments, or control oscillations in the boundary region, this embodiment generates cross-partition coupling control constraints based on the deviation benchmark difference, coupling weight, and actuator response margin: in, Indicates the first The controlled partition and the first The controlled partition in the first The cross-zone coupling control constraint for each control cycle is derived from the zone coupling edge, the response margin of the actuators at both ends, and the difference in the deviation reference value at both ends, and the value ranges from 0 to 1. Indicates the first The controlled partition and the first The coupling weight between the controlled partitions is set to 0.2 to 0.8 in this embodiment. It is derived from the calibration configuration of the proportion of the influence of the output change of adjacent partitions on the sampled value during the initialization of the control system. This setting distinguishes between strongly coupled partitions and weakly coupled partitions. Indicates the first The controlled partition in the first The actuator response margin for each control cycle is derived from the actuator control boundary and the execution output feedback value, and is normalized to a value between 0 and 1. Indicates the first The controlled partition in the first The actuator response margin for each control cycle is derived from the actuator control boundary and the execution output feedback value, and is normalized to a value between 0 and 1. Indicates the first The controlled partition in the first The deviation benchmark of the controlled variable in each control cycle is derived from the aforementioned deviation benchmark calculation results. Indicates the first The controlled partition in the first The deviation benchmark of the controlled variable in each control cycle is derived from the aforementioned deviation benchmark calculation results. This represents the normalized upper limit of the deviation benchmark difference, which is set to 0.5 to 1 in this embodiment, derived from the allowable deviation range of the partitioned controlled variable deviation benchmark. Use 0.5 when the value is less than 0.5 to avoid the denominator being too small. This indicates the controlled partition numbers at both ends of the partition coupling edge, and .

[0050] In the formula results, When writing cross-partition coupling control constraints, if the control directions are consistent, the constraints serve as soft boundaries for subsequent consistency pruning. If the control directions are opposite, the constraints serve as hard boundaries for the difference in control increments between adjacent partitions.

[0051] Furthermore, when writing the control increment difference limit into the cross-zone coupled control constraint, the control system simultaneously records the zone coupling edge number, the controlled zone numbers at both ends, control direction consistency, allowable control increment difference, and actuator response margin. If the control directions of the controlled zones at both ends are opposite and the actuator response margin at either end is less than 0.1, the allowable control increment difference is limited to no more than 0.05. This threshold is derived from the safety margin setting when the actuator approaches the control boundary. This setting avoids forcing a large adjacent control difference when one actuator is already close to saturation. If the actuator response margins at both ends are both greater than 0.5, the allowable control increment difference is determined according to the aforementioned formula to retain the necessary independent adjustment capability of the zones.

[0052] Furthermore, the output of the cross-zone coupling control constraint is used to constrain the candidate control increments of the zones in S3, ensuring that the difference in control increments between adjacent controlled zones does not exceed the boundary calculated by the control system based on the coupling edges. This constraint does not directly change the zone's target variable value, nor does it directly change the actuator's current control command; instead, it serves as a boundary condition before the control increment is generated and enters the consistency pruning process. If a controlled zone does not have adjacent coupling edges, the cross-zone coupling control constraint for that zone is set to 1, indicating that no cross-zone control increment difference limit is applied.

[0053] It should be noted that this step is not a simple difference between the target value and the sampled value, but rather a time window compensation amount is introduced when the confidence of the operating condition switching is not confirmed, and a cross-zone coupling control constraint amount is established between adjacent zones. This processing method enables the control deviation to reflect both the target tracking requirements of the current zone and the coupling boundary between adjacent control loops, overcoming the problem that independent adjustment of each zone in multi-zone lighting control can easily lead to boundary oscillations and mutual interference.

[0054] S3: Generate actuator control increments based on the deviation benchmark of the controlled variable in the partition and the cross-partition coupled control constraints, and update the operating condition fingerprint using the actuator feedback sequence.

[0055] The process of generating actuator control increments includes: converting the deviation benchmark of the controlled variable in a partition into a candidate control increment for that partition; performing consistency trimming on the candidate control increments of adjacent controlled partitions according to the cross-partition coupling control constraints; ensuring that the difference in control increments between adjacent controlled partitions meets the control boundary defined by the cross-partition coupling control constraints; using the trimmed candidate control increments and actuator response margins as inputs to allocate actuator control increments to the actuators in the corresponding controlled partitions; superimposing the actuator control increments onto the current control command of the actuator to form a target control command; outputting the target control command to the corresponding actuator; and writing the target control command into the actuator feedback sequence of the next control cycle.

[0056] Furthermore, the zonal candidate control increment is the initial control change generated by the controlled zonal based on the deviation benchmark of the controlled variable within the current control cycle. Its application is to the zonal control loop, not to individual luminaire structural components. The positive or negative direction of the zonal candidate control increment is determined by the positive or negative direction of the deviation benchmark of the controlled variable. When the deviation benchmark is positive, the candidate control increment is positive, indicating that the control output of the corresponding controlled zonal needs to be increased. When the deviation benchmark is negative, the candidate control increment is negative, indicating that the control output of the corresponding controlled zonal needs to be decreased. The maximum candidate control increment per cycle is set to 0.03 to 0.15, derived from an engineering trade-off between smooth adjustment and control response speed for the lighting control object. This setting avoids abrupt changes in the target control command.

[0057] Furthermore, considering that the controlled variable deviation benchmark and the actuator control command are at different control scales, this embodiment converts the partitioned controlled variable deviation benchmark into a partitioned candidate control increment according to the partitioned control gain, and limits the single-cycle control change: in, Indicates the first The controlled partition in the first The partition candidate control increment for each control cycle is derived from the partition controlled variable deviation baseline and the partition control gain, and its value range is [value range missing]. to . Indicates the first The control gain of each controlled zone is set to 0.2 to 0.8 in this embodiment. It is derived from the response ratio of the controlled variable caused by the change in unit control output during the calibration phase of the control system. This setting makes the control increment of different zones match their response sensitivity. Indicates the first The controlled partition in the first The deviation benchmark of the controlled variable in each control cycle is derived from the output of S2. Indicates the first The maximum candidate control increment per cycle for each controlled partition is set to 0.03 to 0.15 in this embodiment, derived from the engineering preset boundaries of the control system for smooth adjustment and response speed. In the formula results, As input to the consistency trimming, it is not directly output to the executor.

[0058] Furthermore, consistent pruning is used to handle situations where the difference in candidate control increments between adjacent controlled partitions exceeds the cross-partition coupling control constraint. Specifically, for each partition coupling edge, the absolute value of the difference between the candidate control increments of the two partitions is compared with the corresponding cross-partition coupling control constraint. When the absolute value of the difference does not exceed the constraint, the candidate control increments of the two partitions are kept constant. When the absolute value of the difference exceeds the constraint, the excess portion is amortized and reduced according to the response margin of the actuators at both ends, so that the pruned control increment difference satisfies the constraint boundary. If the response margin of the actuators at both ends is 0, the pruned candidate control increments of the two partitions are set to 0 to avoid outputting unexecutable instructions.

[0059] Furthermore, the pruned partition candidate control increments are used to allocate actuator control increments to the actuators within the corresponding controlled partitions. When multiple actuators exist within a controlled partition, the allocation rule uses the actuator response margin as the weight; actuators with larger response margins receive a larger share of control increments, while actuators with response margins below 0.05 are no longer allocated increments in the same direction. The boundary of actuator response margins below 0.05 originates from the safety margin setting of the actuator control range. This setting prevents the continued addition of target control commands when the actuator is close to saturation. If there is only one actuator in the same controlled partition, the pruned partition candidate control increments are directly used as the control increments for that actuator.

[0060] Furthermore, when the actuator control increment is superimposed onto the current actuator control command, the controller imposes boundary restrictions on the execution of the target control command. The normalized value of the target control command ranges from 0 to 1; it is set to 0 when the superimposed result is less than 0, and to 1 when the superimposed result is greater than 1. The boundary is derived from the actuator's allowable control range, ensuring that all target control commands output to the actuator are within the executable range. After the target control command is output, the controller writes it into the actuator feedback sequence of the next control cycle, using it as a comparison benchmark for the actuator control command readback value of the next cycle, thus forming a closed loop of sampling, calculation, output, and feedback.

[0061] After the target control command is output, new execution output feedback values ​​and new controlled variable sample values ​​are received. Based on the target control command, the new execution output feedback values, and the new controlled variable sample values, the execution residual is calculated. The execution residual is used to characterize the degree of deviation between the expected response corresponding to the target control command and the execution output feedback and the actual changes of the controlled variable. The execution residual is written back to the actuator response hysteresis marker and the interval change synchronization relationship in the operating condition fingerprint. When the execution residual is inconsistent with the switching direction corresponding to the operating condition switching confidence value, the operating condition switching evidence is reduced. When the execution residual is consistent with the switching direction corresponding to the operating condition switching confidence value, the operating condition target variable trajectory is retained.

[0062] Furthermore, the execution residual is a feedback quantity used to evaluate whether the control action was executed and whether it caused the expected change in the controlled variable after the target control command was output. The inputs to the execution residual include the target control command, the new execution output feedback value, and the new sampled value of the controlled variable. The new execution output feedback value is returned by the actuator feedback interface after the target control command is output, and the new sampled value of the controlled variable is acquired by the controlled zone sensor in the next sampling cycle. The execution residual ranges from 0 to 1; a larger value indicates a greater deviation between the expected response corresponding to the target control command and the actual feedback.

[0063] Furthermore, considering that comparing only the target control command and the execution output feedback cannot reflect the actual changes in the controlled variable, and comparing only the sampled changes cannot determine whether the actuator responded according to the command, this embodiment uses both the execution output deviation and the deviation of the controlled variable change to calculate the execution residual: in, Indicates the first The execution residual of each controlled partition after the target control command is output comes from the target control command, the new execution output feedback value, and the new controlled variable sample value, and the value is truncated to 0 to 1. The deviation weight of the execution output is set to 0.45 to 0.65 in this embodiment. It is derived from the preset value of the contribution ratio of the execution feedback in the evaluation of the closed-loop control residual. This setting takes into account both the actuator response and the response of the controlled variable. Indicates the first The controlled partition in the first The target control command formed in each control cycle is derived from the superposition of the current control command of the actuator and the control increment of the actuator, and the value range is 0 to 1. Indicates the first The execution output feedback value received by each controlled partition in the next control cycle comes from the actuator feedback sequence and is normalized to a value between 0 and 1. Indicates the first The new controlled variable sample values ​​received by each controlled partition in the next control cycle are derived from the controlled variable sampling sequence and are normalized to a value between 0 and 1. Indicates the first The sampled values ​​of the controlled variables for each controlled partition in the current control period are derived from the sampled sequence of the controlled variables. Indicates the first The control response coefficient of each controlled zone is set to 0.1 to 0.9 in this embodiment, which is derived from the preset value of the response ratio between the control increment and the change of the controlled variable during the calibration phase of the control system. Indicates the first The controlled partition in the first The partition candidate control increment after consistency pruning for each control cycle originates from the consistency pruning result. In the formula result, Used to update the actuator response hysteresis flag and interval change synchronization relationship in the operating condition fingerprint, and to determine whether to reduce the evidence for operating condition switching.

[0064] Furthermore, when writing the execution residual back to the operating condition fingerprint, if the execution residual is greater than the residual threshold, the actuator response hysteresis flag of the corresponding controlled partition is set to 1. If the execution residual is continuously lower than the residual recovery threshold for a recovery hold period, the actuator response hysteresis flag is set to 0. In this embodiment, the residual threshold is set to 0.12 to 0.25, the residual recovery threshold is set to 0.05 to 0.1, and the recovery hold period is set to 2 to 5 control cycles. These values ​​are derived from the actuator feedback error range, the controlled variable sampling noise, and the shortest stable identification period of the control system. This setting avoids frequent switching of the hysteresis flag due to a single abnormal feedback.

[0065] Furthermore, when writing the execution residuals back to the interval change synchronization relationship, the controller compares the magnitude of the execution residuals of adjacent controlled partitions, the direction of change of the controlled variables, and the direction of the trimmed partition candidate control increment. If the execution residuals of adjacent partitions are all below the residual threshold and the direction of change of the controlled variables is consistent with the direction of the control increment, the corresponding interval change synchronization relationship is increased. If the execution residual at one end is higher than the residual threshold while the other end is lower than the residual recovery threshold, or the direction of change of the controlled variables at both ends is opposite, the corresponding interval change synchronization relationship is decreased. The updated interval change synchronization relationship continues to maintain a value range of 0 to 1 and enters the working condition switching evidence calculation in S1 in the next control cycle.

[0066] Furthermore, when the execution residual and the switching direction corresponding to the operating condition switching confidence value are inconsistent, it indicates that the target control command output by the control system has not generated a feedback response matching the operating condition switching direction. In this case, the operating condition switching evidence is reduced. The reduction rule is to multiply the current operating condition switching evidence by a decay coefficient of 0.6 to 0.85. The decay coefficient comes from the preset value of the suppression strength of the control system for erroneous switching confirmation. This setting prevents erroneous operating condition switching from causing the target variable trajectory to deviate continuously. If the execution residual and the switching direction corresponding to the operating condition switching confidence value are consistent, the operating condition target variable trajectory is retained, and the next control cycle is allowed to continue generating partition target variable values ​​according to this trajectory.

[0067] It should be noted that this step, through a closed-loop link of deviation benchmark—candidate control increment—cross-zone consistency pruning—actuator increment allocation—target control command—execution residual writeback, enables the execution result to correct the operating condition fingerprint in reverse. This method is not simply issuing lighting control commands, but rather combining control increment generation and feedback state update into a continuously iterative automatic control process, which can reduce the risk of control oscillation caused by insufficient actuator margin, coupling between adjacent zones, and misjudgment of operating conditions in multi-zone control.

[0068] Example 2, an embodiment of the present invention, provides an intelligent lighting adaptive control system for multiple scenarios, including a fingerprint switching module, a deviation coupling module, and an incremental write-back module. The fingerprint switching module is used to acquire the controlled variable sampling sequence and actuator feedback sequence of the controlled partition, and generate the operating condition fingerprint and operating condition switching confidence value. The deviation coupling module is used to construct the deviation benchmark of the controlled variable in each partition based on the operating condition fingerprint and the operating condition switching confidence value, and to generate cross-partition coupled control constraints. The incremental write-back module generates actuator control increments based on the deviation benchmark of the controlled variables in each partition and the cross-partition coupled control constraints, and updates the operating condition fingerprint using the actuator feedback sequence.

Claims

1. A smart lighting adaptive control method for multiple scenarios, characterized in that, include: Obtain the controlled variable sampling sequence and actuator feedback sequence of the controlled partition, and generate the operating condition fingerprint and operating condition switching confidence value; Based on the operating condition fingerprint and the operating condition switching confidence, a deviation benchmark for the controlled variable in each partition is constructed, and cross-partition coupled control constraints are generated. The actuator control increment is generated based on the deviation benchmark of the controlled variable in the partition and the cross-partition coupled control constraint, and the operating condition fingerprint is updated using the actuator feedback sequence.

2. The adaptive control method for intelligent lighting fixtures oriented towards multiple scenarios as described in claim 1, characterized in that: The operating condition fingerprint includes, The controlled variable sample values ​​are received according to the controlled partition identifier, and a controlled variable sampling sequence is formed according to the sampling time sequence; Receive actuator control command readback values ​​and execution output feedback values ​​according to the actuator identifier, and form an actuator feedback sequence according to the control cycle; Perform time-series smoothing on the sampled sequence of the controlled variable and determine the direction of change of the controlled variable; The actuator response hysteresis flag is determined based on the actuator control command readback value and the execution output feedback value. Write the controlled partition identifier, the direction of change of the controlled variable, and the actuator response lag flag into the operating condition fingerprint.

3. The adaptive control method for intelligent lighting fixtures oriented towards multiple scenarios as described in claim 2, characterized in that: The confidence value of the working condition switching include, Within a continuous control cycle, compare the direction of change of the controlled variable, the actuator response lag marker, and the interval change synchronization relationship in the fingerprints of adjacent operating conditions. The interval change synchronization relationship characterizes the degree of consistency between the change trend of the controlled variable and the response state of the actuator between adjacent controlled partitions; When the direction of change of the controlled variable is inconsistent with the actuator response lag mark, the corresponding control period is marked as the external disturbance period; Evidence of cumulative operating condition switching in control cycles that were never marked as external disturbance cycles; When the evidence of the change of operating condition meets the confirmation conditions for the change and continues to meet the preset retention period, a confidence value for the change of operating condition is generated. When the evidence for the change of operating conditions does not meet the confirmation conditions for the change, the confidence level of the change of operating conditions in the previous control cycle is maintained.

4. The intelligent lighting adaptive control method for multiple scenarios as described in claim 3, characterized in that: The construction of the partitioned controlled variable deviation benchmark includes, Match the trajectory of target variables based on the working condition fingerprint; When the confidence value of the operating condition switching meets the switching confirmation condition, the partition target variable value is generated according to the operating condition target variable trajectory. When the confidence value of the operating condition switching does not meet the switching confirmation condition, the partition target variable value of the previous control cycle is used, and the time window compensation value is generated according to the change of the controlled variable sampling sequence within the sliding time window. The partition target variable value is superimposed with the time window compensation amount, and the difference is calculated with the current sampled value of the controlled variable sampling sequence to obtain the partition controlled variable deviation benchmark. The controlled variable deviation baseline for a zone represents the control deviation between the target variable value and the actual sampled value within the controlled zone after time window compensation.

5. The intelligent lighting adaptive control method for multiple scenarios as described in claim 4, characterized in that: The generated cross-partition coupling control constraint quantities include: Establish partition coupling edges based on the adjacency or coupling influence relationships of the controlled partitions; The partition coupling edge characterizes the mutual influence relationship between adjacent controlled partitions due to changes in controlled variables or actuator outputs; For each partition coupling edge, read the deviation benchmark of the partition controlled variable of the two controlled partitions and calculate the difference of the deviation benchmark of the two controlled partitions. The actuator response margin of the controlled zones at both ends is determined based on the actuator control command readback value and the execution output feedback value; The consistency of the control direction of the two controlled zones is determined based on the deviation benchmark difference. When the control directions of the two controlled zones are opposite, the control increment difference between the two controlled zones is limited by the actuator response margin. Write the control increment difference constraint result into the cross-partition coupled control constraint.

6. The intelligent lighting adaptive control method for multiple scenarios as described in claim 5, characterized in that: The generated actuator control increment includes, Convert the partitioned controlled variable deviation baseline into a partitioned candidate control increment; The candidate control increments of adjacent controlled partitions are uniformly pruned according to the cross-partition coupling control constraints. Consistent pruning ensures that the difference in control increments between adjacent controlled partitions meets the control boundary defined by the cross-partition coupling control constraint. Using the trimmed partition candidate control increment and actuator response margin as input, the actuator control increment is assigned to the actuators in the corresponding controlled partition; The actuator control increment is superimposed on the current actuator control command to form the target control command; The target control command is output to the corresponding actuator and written into the actuator feedback sequence of the next control cycle.

7. The intelligent lighting adaptive control method for multiple scenarios as described in claim 6, characterized in that: The updated operating status fingerprint includes, After the target control command is output, new execution output feedback values ​​and new sampled values ​​of the controlled variable are received; Calculate the execution residual based on the target control command, the new execution output feedback value, and the new sampled value of the controlled variable; The execution residual is used to characterize the degree of deviation between the expected response to the target control command and the execution output feedback and the actual changes of the controlled variable. Write the execution residual back to the actuator response hysteresis marker and interval change synchronization relationship in the operating condition fingerprint; When the execution residual is inconsistent with the switching direction corresponding to the operating condition switching confidence value, the operating condition switching evidence is reduced. When the execution residual is consistent with the switching direction corresponding to the operating condition switching confidence value, the trajectory of the operating condition target variable is retained.

8. A multi-scenario intelligent lighting adaptive control system, employing the multi-scenario intelligent lighting adaptive control method as described in any one of claims 1 to 7, characterized in that: Includes a fingerprint switching module, a deviation coupling module, and an incremental write-back module; The fingerprint switching module is used to obtain the controlled variable sampling sequence and actuator feedback sequence of the controlled partition, and generate the working condition fingerprint and the working condition switching confidence value. The deviation coupling module is used to construct the deviation benchmark of the controlled variable in the partition based on the operating condition fingerprint and the operating condition switching confidence value, and to generate cross-partition coupled control constraint quantities. The incremental write-back module generates actuator control increments based on the deviation benchmark of the controlled variable in the partition and the cross-partition coupled control constraints, and updates the operating condition fingerprint using the actuator feedback sequence.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent lighting adaptive control method for multiple scenarios as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent lighting adaptive control method for multiple scenarios as described in any one of claims 1 to 7.

Images