A smart multi-lamp management and control system and method

Abstract

This invention discloses an intelligent multi-lamp management and control system and method, relating to the field of multi-lamp adaptive control technology. The method includes acquiring historical control command events and constructing a chain-like node information reception density sequence, generating multiple sets of second density schemes, and mapping them to coverage areas and trigger windows for control link execution. It also involves extracting execution features from the chain-like multi-lamp control process and iteratively generating reward values ​​based on Q-values. A density-reward curve is constructed through indexing and global comparison and filtering are performed. This achieves coordinated optimization and optimal density sequence selection for high-density overload and low-density discontinuity problems in chain-like multi-lamp control, solving the command overload conflict caused by high density and the brightness discontinuity and synchronization discontinuity problems caused by low density in multi-lamp chain control.

Description

Technical Field

[0001] This invention relates to the field of multi-lamp adaptive control technology, and more specifically, to an intelligent multi-lamp management and control system and method. Background Technology

[0002] Existing intelligent multi-lamp management and control methods are typically based on a grouped chain topology. This involves a gateway uniformly issuing control commands to the head node of each lamp group, which then forwards them hierarchically along the chain to subsequent nodes, achieving coordinated on / off operation and brightness adjustment control for multiple lamps. In this type of method, the system usually relies on fixed communication relationships and sequential forwarding mechanisms between nodes to disseminate control information, combined with a status feedback mechanism for monitoring operational status, thus forming a hierarchical control architecture of "gateway—head node—subsequent nodes".

[0003] In the aforementioned chain-like control system, the information reception density of each light node is used to characterize the frequency with which each light node receives control commands per unit time. Essentially, it reflects the propagation intensity of control information and the node response rhythm within the chain topology. This density not only affects the execution rhythm of individual nodes but also propagates and amplifies or attenuates along the chain structure, thus directly impacting the synchronization and stability of the overall multi-light system.

[0004] The density of information received by light nodes has a direct structural impact on intelligent multi-light management and control methods. On the one hand, high-density conditions can improve the continuity of control updates and fine-grained response capabilities; on the other hand, low-density conditions can reduce node communication load and improve system stability. However, both high and low densities can introduce different types of system biases during chain propagation, thus constraining the overall control effect.

[0005] Under low-density information reception at light nodes, the number of control commands received per unit time is less, resulting in lower system communication load and reduced node forwarding pressure, thus exhibiting higher operational stability and lower link congestion risk. However, due to the long control update interval, subsequent nodes are prone to segmented or stepped changes in state, leading to discontinuous brightness transitions in multi-light areas and creating a noticeable visual discontinuity effect.

[0006] Under conditions of high-density information reception at light nodes, the frequency of control commands received by nodes per unit time is relatively high. This allows control adjustments to be made continuously in small steps, thereby improving the smoothness and responsiveness of brightness adjustment and giving the multi-light system a stronger following capability in dynamically changing scenarios. However, this state also leads to an increase in the frequency of command superposition during chain propagation. Intermediate nodes are prone to increased information buffering and forwarding pressure, and may even experience competition between old and new commands, resulting in local overload and execution conflicts.

[0007] In existing grouped chain topology control methods, only reliable transmission of control commands and synchronization of node states are usually considered. However, there is a lack of continuous adjustment and structural mapping mechanism design for the information receiving density of lamp nodes during chain propagation. In particular, the propagation impact of high density and low density at different topological locations is not uniformly modeled and coordinated, making it difficult for the system to balance smoothness and stability in complex dynamic scenarios.

[0008] Furthermore, existing technologies often lack a global evaluation mechanism for density changes when dealing with different node information reception densities. They cannot adaptively optimize and select the density distribution based on the chain propagation characteristics, which can easily lead to problems such as local overload, accumulation of synchronization errors, and inconsistent brightness during multi-lamp control, affecting overall control consistency and execution efficiency.

[0009] Therefore, existing technologies lack a unified chain propagation adjustment mechanism and optimization selection method for the high and low states of light node information reception density. It is difficult to ensure system stability while taking into account the continuity and precision of control response. As a result, intelligent multi-light management and control methods have technical defects such as unstable control effect, uneven regional brightness, and unreasonable link load distribution in complex scenarios.

[0010] To address the above problems, this invention proposes a solution. Summary of the Invention

[0011] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an intelligent multi-lamp management and control system and method. By constructing a chain-propagation time-aligned lamp node information reception density modeling and a multi-group density sequence reward evaluation and screening mechanism, the system solves the problems of instruction overload conflict caused by high density and brightness discontinuity and synchronization discontinuity caused by low density in multi-lamp chain control.

[0012] To achieve the above objectives, the present invention provides the following technical solution: A smart multi-lamp management and control method includes the following steps: acquiring a sequence of control command events within a historical unit of time, rearranging it according to gateway—first node—successor node, and dividing the received segments according to time windows, counting the number of command arrivals and interval compression of each segment and superimposing them to obtain the first lamp node information received density sequence; performing several random continuous compression and stretching processes on the time window intensity of the first lamp node information received density sequence, and rearranging and mapping it to the chain node positions to form multiple sets of second lamp node information received density sequences; based on the segmented maintenance mapping of chain propagation time alignment, converting the node information received density sequence into the propagation coverage area from the first node to the successor node and the reception trigger window of the successor node; The information reception density sequence of each second light node is applied to the grouped chain topology. The gateway connects to the first node, and the first node connects to the subsequent nodes in sequence via links. Based on the coverage area and trigger window, hierarchical forwarding and reception control is performed to complete the multi-light activation and deactivation. Features are extracted for each multi-light control, and the corresponding multi-light control reward value is generated by iteratively updating based on the Q value. The information reception density sequence of the second light node is compared hierarchically according to the time window value to generate an index. A curve is fitted with the index as the horizontal axis and the multi-light control reward value as the vertical axis. The information reception density sequence of the second light node corresponding to the maximum multi-light control reward value is selected. If multiple values ​​exist, the one with the smallest window change is selected as the optimal node information reception density sequence and applied to the multi-light control.

[0013] An intelligent multi-lamp management and control system includes an instruction reconstruction module, a density perturbation module, a propagation mapping module, a chain execution module, a feedback evaluation module, and a curve optimization module. The instruction reconstruction module acquires the sequence of control instruction events within a historical unit of time, rearranges it according to gateway-first node-successor node, divides the received segments according to time windows, and counts and superimposes the arrival times and interval compression amounts of each segment to obtain the first lamp node information received density sequence. The density perturbation module performs several random continuous compression and stretching processes on the time window intensity of the first lamp node information received density sequence, and rearranges and maps it to the chain node positions to form multiple sets of second lamp node information received density sequences. The propagation mapping module converts the node information received density sequence into a segmented, time-aligned mapping based on chain propagation to the first node to the successor node. The system includes: a point propagation coverage area and a subsequent node reception trigger window; a chain execution module, which applies the information reception density sequence of each second light node to the grouped chain topology, with the gateway connected to the first node, and the first node sequentially connected to the subsequent nodes via links; a feedback evaluation module, which extracts features from each multi-light control and iteratively updates the corresponding multi-light control reward value based on the Q value; and a curve optimization module, which compares the information reception density sequence of the second light node step by step according to the time window value to generate an index, fits a curve with the index as the horizontal axis and the multi-light control reward value as the vertical axis, selects the second light node information reception density sequence corresponding to the maximum multi-light control reward value, and if multiple sequences exist, selects the one with the smallest window change as the optimal node information reception density sequence and applies it to the multi-light control.

[0014] An electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform an intelligent multi-lamp management control method.

[0015] A computer-readable storage medium storing a computer program that, when executed by a processor, implements an intelligent multi-lamp management and control method.

[0016] In the embodiments provided by this invention, it should be understood that the disclosed system can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation. The technical effects and advantages of the intelligent multi-lamp management control system and method of this invention are as follows: 1. This invention constructs a node information reception density sequence generation mechanism based on chain-like propagation timing alignment. It uniformly rearranges and windowes historical control command events within the gateway-first node-successor node chain topology, and introduces a segmented mapping relationship between coverage area and reception trigger window. This transforms the originally discrete command reception behavior into a continuous control expression with propagation structure constraints, thereby achieving a structured characterization and adjustable expression of the control information propagation process in a multi-lamp chain system. This technology enables each node to maintain a unified reference benchmark for reception rhythm at different time scales, improving the consistency of information synchronization during chain control, reducing control asynchrony problems caused by time offsets, and thus enhancing the overall stability and continuity of the multi-lamp system's scheduling.

[0017] 2. This invention performs chain-like execution simulation evaluation on multiple sets of second-lamp node information reception density sequences, and generates multi-lamp control reward values ​​by iteratively updating the Q-value. A control reward curve is constructed on the index dimension, enabling a global comparison and optimal selection mechanism for different density distribution schemes. This establishes a quantifiable balance between high-density response smoothness and low-density system stability. This technology effectively suppresses link overload and command conflicts caused by high density, while avoiding brightness staggering and regional discontinuities caused by low density. The final selected node information reception density sequence ensures chain-like propagation stability while possessing superior control continuity and system resource utilization efficiency, thus significantly improving the overall operational quality of multi-lamp collaborative control. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating an intelligent multi-lamp management and control method according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] Example 1, Figure 1 The present invention provides an intelligent multi-lamp management and control method, comprising the following steps: S1: Obtain the sequence of control command events within a historical unit of time, rearrange them according to gateway—first node—successor node, divide the received segments according to time window, count the number of command arrivals and interval compression of each segment and superimpose them to obtain the first light node information received density sequence.

[0021] In this embodiment, the process of obtaining the control command event sequence within a historical unit of time, rearranging it according to gateway—first node—successor node, dividing the received segments according to time windows, and counting and superimposing the number of command arrivals and interval compression of each segment to obtain the first light node information received density sequence is as follows: Obtain the sequence of control command events received by each light node in the grouped chain topology within a unit time, and divide the received events of each light node into a set of continuous time window segments according to a fixed-length time window. For each time window segment, the number of instruction arrivals is extracted, and the average convergence amplitude of the time interval between two adjacent instruction arrivals within each window is calculated. The average convergence amplitude is the ratio of the sum of the absolute values ​​of the differences between adjacent time intervals to the number of intervals. The number of instruction arrivals is scaled by the maximum and minimum value range, and the average convergence amplitude is shifted by the mean center, so that the two are mapped to the same numerical interval endpoint range. Within the same time window, the difference between the number of arrivals of the mapped instructions and the average convergence amplitude is statistically analyzed item by item, and the difference between each time window is accumulated and summed segment by segment to obtain the received strength value corresponding to the time window. The average of the received strength values ​​corresponding to each time window is used to form the first light node information received density for the corresponding light node. Following the chained order of gateway—first node—successor node, the first light node information reception density of each light node is arranged sequentially to form the first light node information reception density sequence.

[0022] In this embodiment, the sequence of control instruction events per unit time is a set of all instruction trigger records recorded during the process of the gateway issuing instructions to the first node and the first node forwarding instructions to subsequent nodes in the grouped chain topology within a preset time range. The corresponding sets are arranged in chronological order of occurrence. Each record contains at least the instruction generating node, the receiving node, and the corresponding occurrence time, which is used to reflect the propagation trajectory and distribution of instructions in the network during the chain control process.

[0023] In this embodiment, a time window is a discrete time period formed by dividing a continuous time axis into equal intervals according to a fixed duration. Each time window is used to carry all instruction receiving events that occur within the corresponding time range, so that the control behavior in different time periods can be segmented, statistically analyzed, and compared, thereby avoiding the statistical aliasing problem caused by a single continuous time series.

[0024] In this embodiment, the received segment is a set of light node instruction reception records collected within each time window. The corresponding set consists of all arriving instruction events within the corresponding time window and is arranged in the order of occurrence. It is used to describe the actual control reception behavior process of a single light node within the corresponding time window.

[0025] In this embodiment, the number of command arrivals is the statistical value of the number of control commands actually received by a certain light node within a single time window. The corresponding value is obtained by counting the event records in the received segment within the corresponding time window one by one, and is used to characterize the control interaction frequency intensity of the corresponding node within the corresponding time period.

[0026] In this embodiment, the interval compression amount describes the degree of shortening of the time interval between two adjacent instructions relative to the historical stable interval benchmark. The benchmark is determined by the average level of the interval between adjacent instructions at the corresponding node in the historical time window. By comparing the degree of shortening of the interval change in the current window relative to the historical benchmark, the compression amount is accumulated to characterize the density of instructions arriving in a short period of time.

[0027] In this embodiment, the average convergence amplitude is the average level of the absolute value of the difference between the time intervals of all adjacent instructions within the same time window. The corresponding value is formed by statistically analyzing the variation amplitude between consecutive instruction intervals within the corresponding window and taking the equilibrium value, which is used to characterize the stability and convergence trend of the instruction interval within the corresponding window.

[0028] In this embodiment, the maximum and minimum value range scaling process uses the minimum and maximum values ​​of the number of instruction arrivals in the entire history window as boundaries, and maps the number of instruction arrivals in the current window to a unified numerical range according to a proportional relationship, so that the instruction frequencies between different nodes have a unified comparison scale and avoid the incomparability problem caused by the difference in absolute values ​​between different nodes.

[0029] In this embodiment, the mean center offset processing uses the average convergence magnitude in the historical window as the reference center and re-expresses the deviation of the average convergence magnitude of the current window from the corresponding reference center. This allows the convergence changes between different time windows to be compared and analyzed under a unified reference center, thereby eliminating the impact of time scale fluctuations.

[0030] In this embodiment, the endpoint range of the same numerical interval is a standardized mapping interval used to uniformly constrain the expression range of different feature quantities. The corresponding interval serves as a unified expression space after the number of instruction arrivals and the average convergence amplitude are jointly mapped, enabling two types of features with different physical meanings to participate in subsequent difference statistics at the same scale.

[0031] In this embodiment, the item-by-item difference statistics are performed by comparing the number of times the mapped instructions arrive with the average convergence amplitude element by element within the same time window, and recording the numerical difference at each corresponding position to express the degree of deviation between the two types of control behavior characteristics within the same time window.

[0032] In this embodiment, the segment-by-segment cumulative summation is to continuously superimpose the difference sequence obtained in each time window according to the time window order, so that the deviation of each window can form a continuous growth or fluctuation expression on the overall time axis, thereby forming an overall characterization of the node reception intensity.

[0033] In this embodiment, the received strength value is a single numerical expression obtained after difference statistics and accumulation processing for each time window, which is used to represent the unified result of the overall tightness and fluctuation of the light node control command reception within the corresponding time window.

[0034] In this embodiment, the first light node information reception density is the average expression of the reception intensity value of a single light node in all time windows within a unit time range, which is used to describe the overall level of the corresponding node's command reception carrying capacity and control density.

[0035] In this embodiment, the first light node information reception density sequence is an ordered set of values ​​formed by arranging the first light node information reception densities corresponding to each light node in sequence according to the inherent connection order of the gateway, the first node and the successor nodes in the grouped chain topology. This set is used for subsequent chain propagation analysis and control optimization calculations.

[0036] It should be noted that by mapping the number of command arrivals and the amount of interval compression to a unified scale, and by performing difference accumulation processing on the two types of features at the time window level, the originally heterogeneous control event information is transformed into a uniformly comparable expression of received strength, thereby avoiding the problem that a single frequency or single time interval indicator cannot fully reflect the chain propagation load.

[0037] It should be noted that the corresponding method introduces a time window segmentation mechanism and a chain-like sequential arrangement, enabling control command events distributed across different nodes and time scales to be modeled uniformly within the same structural framework, providing a stable input foundation for subsequent control optimization based on propagation coverage and trigger windows.

[0038] It should be noted that the beneficial effect of the corresponding steps is that the traditional single evaluation method that only relies on the number of instructions or the delay is expanded into a composite receiving density expression that considers both the instruction density and the convergence characteristics of interval changes. This allows the load state of the lamp nodes in the chain topology to be more finely characterized, thereby improving the stability and response consistency of subsequent multi-lamp collaborative control.

[0039] S2, the time window intensity of the first lamp node information receiving density sequence is subjected to several random continuous compression and stretching processes, and then rearranged and mapped to the chain node position to form multiple sets of second lamp node information receiving density sequences.

[0040] In this embodiment, the time window intensity of the first light node information reception density sequence is subjected to several random continuous compression and stretching processes, and then rearranged and mapped to the chain node positions to form multiple sets of second light node information reception density sequences, specifically as follows: For each node density value in the received density sequence of the first light node information, extract the corresponding time window intensity array in a fixed order, and divide the corresponding array into a continuous local window sub-segment with a length of three. For each local window segment, a transformation operation is performed. The transformation operation includes shifting the intensity difference between adjacent windows in the segment sequentially backward by one time step to generate a compressed transformation segment, and inserting an equal number of zero compensation terms into the intensity difference between adjacent windows in the segment sequentially to generate a stretched transformation segment. The compression transformation segments and the stretch transformation segments are alternately combined according to the original time window order to form a reconstruction sequence of several time windows corresponding to the nodes; For each time window reconstruction sequence, the topological position index of the gateway-first node-successor node is randomly rearranged so that the intensity of each time window in the same sequence corresponds one-to-one with the corresponding node number, forming a node mapping sequence. Based on the time window strength of the first node in the node mapping sequence, sequential delay alignment is performed on the time window strength of its successor nodes to ensure that the time window strength of all nodes is expressed with the same time step length. The aligned node mapping sequence is output as a set of second lamp node information receiving density sequences, and the alignment is repeated for all reconstructed sequences to form several sets of second lamp node information receiving density sequences.

[0041] In this embodiment, the first light node information reception density sequence is an ordered set of light nodes arranged in the order of gateway, first node and successor node, representing the density of received instructions per unit time. Each element in the sequence corresponds to the overall reception density value of a light node, which is used as the input basis for subsequent time structure perturbation and topology reconstruction.

[0042] In this embodiment, the time window intensity array is a set of values ​​formed by arranging the reception intensity values ​​of each consecutive time window of a single light node in chronological order during the construction of the information reception density of the first light node. The corresponding set is used to express the change process of the reception behavior of the corresponding node in different time segments.

[0043] In this embodiment, the local window segment is a subsequence structure formed by dividing the time window intensity array into segments of three consecutive time windows of fixed length. Each segment contains the received intensity values ​​of three adjacent time windows, which is used to characterize the dynamic changes within a local time range.

[0044] In this embodiment, a local window segment with a continuous length of three is a time segment structure formed by three adjacent time windows as the smallest analysis unit. By using a fixed length constraint, the consistent expression of different nodes on the time scale is ensured, so that local change features can be compared and transformed in a unified structure.

[0045] In this embodiment, the compressed transformation segment is a reconstructed sequence formed by shifting the time position of the intensity difference between adjacent time windows within a local window segment. The corresponding processing method enables the originally dispersed change trend to be expressed in a concentrated manner within a shorter time step, which is used to simulate the situation where instructions are rapidly gathered in chain propagation.

[0046] In this embodiment, the stretching transformation segment is an extended sequence formed by inserting zero-value compensation terms corresponding to the original difference scale between adjacent time windows within the local window segment. The corresponding processing method extends the original change process in the time dimension, which is used to simulate the situation where the instruction propagation interval is enlarged.

[0047] In this embodiment, the zero compensation term is a placeholder value that is artificially inserted between time windows without any change. The corresponding value indicates that there is no new change in received intensity within the corresponding time step. It is used to expand the time scale while maintaining the continuity of the sequence structure, so that different nodes can be aligned in the time structure.

[0048] In this embodiment, the alternating combination is a processing method that alternately arranges the compression transformation segments and the stretch transformation segments according to the original time window order to form a complete sequence. This method is used to construct a composite time structure expression that simultaneously contains high-density concentrated changes and low-density extended changes.

[0049] In this embodiment, the time window reconstruction sequence is a node-level time intensity sequence formed by alternating compression and stretching. The corresponding sequence retains the original node received intensity change trend, while introducing a multi-scale time perturbation structure to generate diverse node received behavior expressions.

[0050] In this embodiment, the chain topology location index is a fixed arrangement number of the gateway, first node, and successor nodes in the grouped chain structure, which is used to specify the propagation order between nodes so that the time window intensity can be rebound according to the chain structure in the topology dimension.

[0051] In this embodiment, the node mapping sequence is a sequence structure formed by re-corresponding the time window reconstruction sequence according to the chain topological position index. The correspondence structure redistributes the time window intensity of different nodes under a unified topological order, which is used to construct the correspondence relationship between different nodes.

[0052] In this embodiment, random rearrangement is a process of randomly exchanging and recombining the correspondence between time window intensity and node number while maintaining the chain topology index constraint. This process is used to generate various combinations of different receiving density distributions among nodes, thereby increasing the sample diversity for subsequent control optimization.

[0053] In this embodiment, the sequential delay alignment process is based on the time step where the first node's time window intensity is located, and the time window intensity of the subsequent nodes is shifted according to a fixed time step difference, so that all nodes are aligned on the same time step length, thereby ensuring that the time changes between different nodes are comparable.

[0054] In this embodiment, the second light node information receiving density sequence is a set of multiple node-level receiving densities formed after compression and stretching transformation, topology rearrangement and time alignment processing. Each set of sequences corresponds to a possible chain propagation structure configuration, which is used for subsequent control strategy evaluation and optimization selection.

[0055] It should be noted that by alternating compression and stretching of the time window intensity, the original single-time-scale receiver density expression is expanded into a multi-scale dynamic structure, thereby simulating the changes in instruction transmission state under different propagation rhythms in a chain topology.

[0056] It should be noted that by introducing a topological position random rearrangement and time alignment mechanism, the reception density characteristics of the same node can be recombined under different chain structure arrangements, thereby constructing multiple propagation path hypotheses and improving the adaptability of subsequent control strategies to complex chain scenarios.

[0057] It should be noted that the beneficial effect of the corresponding steps is that, through the joint processing of temporal structure perturbation and topology reconstruction, a single historical received density sequence is transformed into multiple sets of sequence inputs with different propagation rhythms and node distribution characteristics, providing a rich candidate space for subsequent optimization and screening based on coverage and trigger window, thereby improving the robustness and adjustment accuracy of multi-lamp chain control.

[0058] S3, based on the segmented preservation mapping of chain propagation time alignment, converts the node information reception density sequence into the propagation coverage of the first node to the subsequent nodes and the reception trigger window of the subsequent nodes.

[0059] In this embodiment, the segmented preservation mapping based on chain-propagation timing alignment converts the node information reception density sequence into the propagation coverage area from the first node to the subsequent nodes and the reception trigger window of the subsequent nodes, specifically as follows: The starting time marker corresponding to the continuous expression sequence of the received intensity per unit time of each light node is set as the relative time when the node accesses the gateway, and the time when the gateway issues the first control command is used as the zero point reference of the global time axis for chain propagation. Starting from the zero point of the global time axis, the time offset of each time window in the continuous expression sequence of the received intensity per unit time of each lamp node is calibrated. The time offset is determined by the time difference between the time when the corresponding node accesses the gateway and the zero point of the global time axis. Subtract the corresponding time offset from the start and end boundaries of each time window in the time window sequence of each light node, so that the time window sequence of all nodes is converted to a unified global time axis scale system. Under a unified global timeline scale system, the time windows of different light nodes are rearranged segment by segment according to the time window start-point alignment rule, so that the time windows within the same time scale range have a consistent start and end boundary expression on the global timeline. On the continuous expression sequence of received intensity per unit time after global time axis alignment, based on the principle of keeping the original time window boundaries unchanged, the continuous expression sequence of each lamp node is divided into segments according to the start and end boundaries of the corresponding time window, forming a segmented sequence that corresponds one-to-one with the original time window. For each hold segment, the segment duration and segment intensity peak are extracted. The segment duration is determined by the duration span of the corresponding time window on the global time axis, and the segment intensity peak is determined by the maximum value in the continuous expression sequence of received intensity per unit time within the corresponding segment. The segment duration is used as a measure of propagation coverage capability, and the segment intensity peak is used as a benchmark for propagation intensity constraints. The propagation coverage capability representation is passed level by level in a chain topology order from gateway to first node to successor node, and at each successor node, the propagation coverage capability representation is updated by decreasing according to the link propagation delay constraint between adjacent nodes, generating a set of continuous coverage nodes on the corresponding chain topology path. Within the holding segment corresponding to each successor node, a symmetrical time expansion interval is constructed on the global time axis, with the time position corresponding to the peak intensity of the segment as the reference point, and the interval is used as the receiving trigger time interval of the corresponding node. By binding and matching the continuous set of covered nodes with the corresponding receiving trigger time intervals of each successor node on the global time axis, a segmented and persistent mapping structure based on chain propagation timing alignment is formed.

[0060] In this embodiment, the continuous expression sequence of received intensity per unit time is a set of received intensity values ​​arranged by time window formed by each light node during the construction of the first light node information received density. The corresponding set records the changes in the strength of instruction received by the node in different time windows in chronological order, which is used as the basic input data for subsequent time alignment and propagation modeling.

[0061] In this embodiment, the start time marker is the time record corresponding to the arrival time of the first instruction for each light node to start participating in the chain control reception. The corresponding time is used to identify the access start point of the corresponding node relative to the overall chain propagation system and serves as a reference benchmark for subsequent time offset calculation.

[0062] In this embodiment, the zero point of the global time axis for chain propagation is the moment when the gateway issues the first control command. The corresponding moment serves as a unified reference starting point for time alignment of all nodes in the entire grouped chain topology, which is used to eliminate time scale deviations caused by differences in access times of different nodes.

[0063] In this embodiment, the time offset is the time difference between the starting time marker of each lamp node and the zero point of the global time axis of chain propagation. The corresponding value is used to uniformly map the received intensity sequences of different nodes to the same time reference system, so that the time windows of all nodes can be aligned and compared on the same time axis.

[0064] In this embodiment, subtracting the corresponding time offset from the start and end boundaries of the time window is a processing method that performs a unified time axis translation on the start and end times of the original time window of each node. This processing transforms the originally independent time window structures of different nodes into a unified expression structure under a shared global time axis.

[0065] In this embodiment, the time window start-point alignment rule is to align and sort the time windows of all nodes according to their start time under the global time axis scale system, so that the receiving behaviors of different nodes occurring within the same time range form a segmented correspondence on the time axis, which is used to ensure that the multi-node propagation process is comparable at the same time scale.

[0066] In this embodiment, the segment sequence is a set of segments formed by dividing the continuous expression sequence of received intensity per unit time according to the principle of keeping the original time window boundary unchanged after global time axis alignment. The corresponding set maintains the original time window division structure and is used to preserve the original time segmentation features of the nodes.

[0067] In this embodiment, the segment duration is the time span of each maintenance segment on the global time axis. The corresponding span is determined by the start and end time difference of the corresponding time window, which is used to characterize the time coverage capability of the corresponding node to continuously participate in control reception within the corresponding time interval.

[0068] In this embodiment, the peak intensity of a segment is the maximum value among all time window received intensity values ​​within each holding segment. The corresponding value is used to represent the intensity level at the moment when the node receives the most instructions within the corresponding time interval, and is used as a reference for propagation capability constraints.

[0069] In this embodiment, the propagation coverage capability is represented by a chain propagation capability index obtained by mapping the segment duration. The corresponding index is used to describe the node range time basis in which the instruction can act continuously in the chain topology, and is used to constrain and control the spatial expansion capability of propagation.

[0070] In this embodiment, the propagation strength constraint benchmark is a constraint index composed of the peak intensity of the segment. The corresponding index is used to limit the maximum allowable received strength level of the control command during propagation, so as to avoid the node from overloading under high-density conditions.

[0071] In this embodiment, the chain-like topology sequential propagation is a process of propagating the coverage capability representation quantity in sequence according to the connection relationship between the gateway, the first node, and the successor node. The corresponding process allows the coverage capability to be passed from the upstream node to the downstream node step by step, forming a propagation constraint structure that decreases layer by layer.

[0072] In this embodiment, the propagation delay constraint is the time difference constraint between two adjacent nodes during the instruction transmission process. The corresponding constraint is used to limit the range of coverage capability at subsequent nodes, so that the propagation capability gradually decreases as the link length increases.

[0073] In this embodiment, the continuous coverage node set is a set of nodes that still meet the coverage conditions during the process of progressively transferring and decreasing the coverage capability. The corresponding set is used to identify the range of chain topology nodes that the current control command can effectively operate on.

[0074] In this embodiment, the receiving trigger time interval is a time range interval formed by taking the time position corresponding to the peak intensity of the segment as the center point and expanding symmetrically according to a fixed time. The corresponding interval is used to define the effective time window for subsequent nodes to execute control actions.

[0075] In this embodiment, the reference point is the time position where the received strength reaches its maximum value within each holding segment. The corresponding position represents the moment when the node response is strongest within the corresponding time window, and is used as the central reference point for triggering window construction.

[0076] In this embodiment, the segmented mapping structure is an overall structural expression formed by binding the set of continuous coverage nodes and the receiving trigger time interval of each successor node one by one on the global time axis. The corresponding structure is used to describe the spatial propagation range and time trigger constraint relationship in the chain topology.

[0077] It should be noted that by introducing a global timeline alignment mechanism, the time window structures that were originally scattered across different nodes are unified under the same time reference system, thereby eliminating the offset effect caused by the difference in node access time on propagation analysis.

[0078] It should be noted that by using the segment duration and segment intensity peak as the propagation capability and constraint benchmarks respectively, the chain propagation process has both coverage control and intensity limitation mechanisms, thereby avoiding overload or fault problems caused by single-dimensional modeling.

[0079] It should be noted that the beneficial effect of the corresponding steps is that, through joint modeling of time alignment, segment preservation, and coverage propagation, the node reception density can be converted into an executable propagation range and trigger window expression, thereby providing a control basis for subsequent multi-lamp chain control that has both time consistency and topological constraints.

[0080] S4. The information reception density sequence of each second light node is applied to the grouped chain topology. The gateway is connected to the first node, and the first node is connected to the successor nodes in sequence through the link. According to the coverage area and the trigger window, the hierarchical forwarding and reception control is performed to complete the multi-light opening and closing.

[0081] In this embodiment, the density sequence of information received by each second light node is applied to a packet chain topology. The gateway is connected to the first node, and the first node is connected to subsequent nodes sequentially via links. Based on the coverage area and trigger window, hierarchical forwarding and receiving control is performed to complete the multi-light activation and deactivation. Specifically: For each second light node information receiving density sequence, a group-level control instruction packet is first generated on the gateway side, and the corresponding propagation coverage range parameter and subsequent node receiving trigger window parameter are written into the instruction packet, while the group identifier and sequence identifier are attached. The gateway sends the instruction packet to the first node of the corresponding light group. After receiving it, the first node performs chain mapping decomposition on the propagation coverage parameters, generates the local coverage segment and the next-hop forwarding segment corresponding to the first node, and splits the received trigger window parameters into time intervals to generate its own execution window and the next-level forwarding window. The first node performs control actions within the trigger time interval according to its own execution window, and initiates a chain-like forwarding process within the coverage area, synchronously sending control commands to the next successor node; After each successor node receives the instruction, it confirms whether it is within the effective coverage range based on its corresponding coverage segment parameters, and determines whether to enter the execution state window based on the received trigger window parameters. When the conditions are met, it performs the corresponding opening / closing or dimming action, and synchronously generates the next hop coverage segment and trigger window. Each successor node repeats the above-mentioned coverage segment inheritance and trigger window inheritance process level by level along the chain topology, so that the propagation coverage range is passed along the link level by level and decays level by level, while the trigger window remains continuously aligned and propagates on the time axis. When the coverage area propagates to the end node or the coverage segment decays to the preset termination condition, forwarding stops, completing the multi-lamp on / off control of the chain control execution process corresponding to the information reception density sequence of a single second lamp node.

[0082] In this embodiment, the group-level control instruction package is a structured control data unit generated by the gateway for the information reception density sequence of a single second light node. It is constructed by uniformly encapsulating the corresponding sequence number, chain topology number, propagation coverage parameters and subsequent node reception trigger window parameters, and adding a group identifier to distinguish different control tasks and a sequence identifier to distinguish different time evolution strategies, so that the corresponding instruction package can uniquely correspond to a multi-light control execution path in the chain network.

[0083] In this embodiment, the propagation coverage parameter is a quantitative expression of the boundary of the number of nodes that can be acted upon step by step from the first node to the subsequent nodes. It is formed by calculating the upper limit of the chain propagation capability corresponding to the distribution of the received intensity per unit time in the information received density sequence of the second light node. Specifically, it is the range of nodes that can continuously participate in control forwarding, which is used to limit the maximum propagation depth of control commands in the chain topology.

[0084] In this embodiment, the subsequent node receives the trigger window parameter, which is a set of time intervals in which the subsequent node is allowed to enter the control execution state. It is formed by expanding the time interval based on the intensity peak distribution position of the second light node information receiving density sequence on the time axis, so that each node responds to the control command only within the corresponding window within the specified time range, thereby avoiding repeated triggering across time windows.

[0085] In this embodiment, the chain mapping is decomposed into a process in which the first node splits the propagation coverage parameters level by level. The splitting method is to divide the overall coverage into a single-node coverage segment and a next-hop forwarding segment according to the node order. The single-node coverage segment is used to determine the range that the current node is allowed to directly act on, and the next-hop forwarding segment is used to determine the remaining range to continue propagating to subsequent nodes.

[0086] In this embodiment, the local coverage segment is the local range of action obtained by each node in the chain topology based on the propagation coverage range parameter. It is determined by the position number of the corresponding node in the link and the coverage range parameter, and is used to limit the number of successor nodes that the corresponding node can directly affect in the current control cycle.

[0087] In this embodiment, the execution window is the time interval of the node's own action determined by the received trigger window parameters. It is formed by extracting the peak corresponding time point from the received trigger window parameters and expanding a fixed time span forward and backward with the corresponding time point as the center, which is used to limit the time range of the node's actual execution of the opening, closing or dimming action.

[0088] In this embodiment, the lower-level forwarding window is the time interval during which a node is allowed to continue sending control commands to the next node after completing its own execution action. It is formed by extending the fixed link propagation delay after the execution window ends, so as to ensure that the chain forwarding is continuous in time but does not conflict with the execution window.

[0089] In this embodiment, the chain-like forwarding process is an execution mechanism in which control commands are transmitted sequentially along a preset unidirectional link after being triggered by the first node. The process is that after the current node meets the coverage segment and trigger window conditions, it sends the same set of level control command packets to the only successor node, and the successor node repeats the corresponding process until the end node.

[0090] In this embodiment, the effective coverage range is the range of conditions under which subsequent nodes determine whether to continue participating in control propagation. The determination method is to compare the current node's location with the boundary of the propagation coverage segment. If the node's sequence number is within the coverage segment, it is allowed to participate in control execution; otherwise, forwarding is stopped.

[0091] In this embodiment, continuous alignment propagation is a propagation method that triggers windows to maintain temporal consistency in a chain topology. It is implemented by each node using the same global time axis as a reference when generating its own execution window and forwarding window, so that the start and end boundaries of the window maintain a relatively synchronous relationship between different nodes.

[0092] It should be noted that this step introduces a dual constraint mechanism of propagation coverage and receiving trigger window during the control execution process, so that the control command no longer depends on a single time trigger, but is simultaneously restricted by the link depth and time window, thus forming a dual adjustment structure of controllable propagation depth and controllable execution timing.

[0093] It should be noted that in the chain topology, this step uses the method of inheriting the coverage segment and trigger window level by level to make the control command gradually attenuate the propagation range and maintain the continuity of time during the transmission process, thereby avoiding the problems of global synchronization loss of control or local node over-response in multi-lamp systems.

[0094] It should be noted that the beneficial effect of this step is that by transforming the second light node information reception density sequence into a combined control mechanism of coverage and trigger window, the multi-light start-up and shutdown process can achieve predictable and constrained step-by-step propagation control in a chain structure, thereby improving the timing consistency and execution stability of the overall system.

[0095] S5 extracts features from each multi-lamp control and generates the corresponding multi-lamp control reward value based on the Q value through iterative updates.

[0096] In this embodiment, the step of extracting features from each multi-light control and generating corresponding multi-light control reward values ​​based on Q-value iterative updates is as follows: Based on the chain control execution process corresponding to the information reception density sequence of each second light node, the sequence of instruction reception times, the sequence of repeated reception times, and the sequence of adjacent instruction arrival time intervals of each node are extracted in the propagation order of gateway, first node and subsequent nodes. The sequence of instruction coverage overlap times and forwarding waiting time of each node in the same time window is also recorded to form node-level propagation execution characteristics. The node-level propagation execution features are merged and spliced ​​in a chain-like topological order to form a chain-like propagation state representation corresponding to the information reception density sequence of the second light node; In the chain propagation state representation, a feature decomposition set is constructed to characterize the differences in receiver density, including a high-density chain propagation feature set and a low-density chain propagation feature set, specifically: For high-density chain propagation feature groups, based on the sequence of instruction reception counts within a unit time window, the length of the continuous interval corresponding to the first threshold multiple of the reception count of each window in three or more consecutive time windows is extracted as the high-density continuous reception interval length, which is used to characterize the execution feature of continuous brightness adjustment and fine response when the reception density is high. Based on the instruction arrival interval sequence within the same time window, the percentage of times the interval value is lower than the second threshold multiple of the historical average interval of the node is statistically analyzed to form the interval compression ratio feature, which is used to characterize the state of rapid and continuous instruction superposition under high density conditions. Based on the overlap sequence, the number of conflicts occurs when the overlap number exceeds the maximum number of conflict-free overlaps in the history of a node within a single time window, forming an instruction conflict frequency feature to characterize the conflicts caused by the superposition of new and old instructions due to high density. Based on the forwarding wait time sequence, the number of intervals in which the wait time continuously increases within three or more consecutive time windows is extracted to form the link wait growth interval feature, which is used to characterize the information overload and forwarding blocking status of intermediate nodes. For low-density chain propagation feature groups, based on the sequence of instruction reception counts within a unit time window, the length of the continuous interval corresponding to the third threshold multiple when the number of receptions in each of three or more consecutive time windows is lower than the average number of receptions in the historical windows of the node is extracted, and used as the length of the low-density continuous sparse interval to characterize the state of reduced node load but insufficient response. Based on the instruction arrival interval sequence, the percentage of times the interval value is higher than 1.8 times the historical average interval of the corresponding node is counted to form a high interval percentage feature, which is used to characterize the sparse instruction propagation and response delay state; based on the brightness change sequence of the successor node, the number of consecutive times when the difference between the brightness change amplitude of two adjacent time windows continuously exceeds the fourth threshold of the historical average change amplitude is counted to form a brightness step jump number feature, which is used to characterize the brightness change as a segmented jump state. Based on the time alignment difference between brightness change sequences of different nodes, the number of asynchronous duration intervals exceeding one time window length is counted to form multi-node tomographic interval features, which are used to characterize the discontinuous brightness state of multi-lamp regions. The high-density chain propagation feature group and the low-density chain propagation feature group are combined in a fixed order to form a chain propagation feature vector corresponding to the information receiving density sequence of each second lamp node. The second light node information receiving density sequence is used as the action input, the chain propagation feature vector is used as the state input, and the continuous offset of brightness change of each node, the number of command conflicts and the link forwarding delay fluctuation during the chain control execution process are used as the feedback sources. The continuous offset of brightness change is the absolute accumulation of the difference between the brightness change amplitude of adjacent time windows and the average of the historical stable change amplitude, which is used to characterize the control smoothness. The number of instruction conflicts is the count of execution conflicts caused by repeated instruction overwriting within the same time window, used to characterize high-density overload penalties; Link forwarding delay fluctuation is the cumulative variance of the instruction arrival time difference between adjacent nodes, used to characterize the low-density propagation fault penalty; Based on the state input, action input, and reward source, the chain control execution process corresponding to the information receiving density sequence of each second light node is updated sequentially to generate the corresponding single multi-light control reward value.

[0097] In this embodiment, the node-level propagation execution feature is a structured description formed by hierarchically recording the behavior of each node during the chain control execution process. The construction method is to continuously record the number of instructions received, the number of repeated receptions, and the time interval between adjacent instructions in each unit time window for each node in the order of gateway, first node, and successor nodes. At the same time, the number of instruction overlaps and the forwarding waiting time in the same time window are superimposed to form a basic feature set that can reflect the real-time reception and forwarding status of the node in the chain propagation.

[0098] In this embodiment, the instruction reception sequence is a sequence formed by arranging the number of control instructions actually received by each node within a unit time window in chronological order. It is generated by counting each control instruction arriving at the node within each time window and recording them sequentially according to the time window order, which is used to characterize the change in the receiving load of the node at different time stages.

[0099] In this embodiment, the repeated reception sequence is a record of the number of times the same node receives the same control instruction or repeated propagation instruction within the same time window. It is formed by identifying instructions with the same sequence number or the same control content and counting the number of times they appear repeatedly, which is used to characterize the degree of instruction redundancy during chain propagation.

[0100] In this embodiment, the instruction arrival time interval sequence is a continuous record consisting of the arrival intervals of two adjacent control instructions on the time axis. It is formed by sequentially differentiating the timestamps of the nodes receiving instructions to obtain the time intervals, and classifying and recording them according to time windows. This is used to characterize the continuity and sparsity of instructions in chain propagation.

[0101] In this embodiment, the instruction overlap count is the number of times that multiple control instructions overlap in their execution time at the node execution level within the same time window. It is formed by comparing the start and end intervals of different instructions within the execution time window, and performing an overlap count when the intervals intersect, which is used to characterize the degree of parallel conflict during instruction execution.

[0102] In this embodiment, the forwarding wait time sequence is a record of the time delay between a node receiving a control command and forwarding it to the next node. It is formed by recording the time difference between each received event and the corresponding forwarding event and arranging them according to time windows, which is used to characterize the processing delay and blocking situation in the link propagation.

[0103] In this embodiment, the chain propagation state is represented as a unified state vector structure formed by splicing all node-level propagation execution features according to the topological order from the gateway to the end node. The construction method is to arrange the number of receptions, repetitions, time intervals, coverage overlaps, and forwarding wait times of each node in the order of nodes, which is used to express the overall operating state of the entire chain system within a single control cycle.

[0104] In this embodiment, the high-density chain propagation feature set is a feature set used to characterize the system behavior when control commands arrive at high frequencies per unit time. It is formed by combining the statistical results of continuous window intervals with a reception frequency higher than the historical average, time segments with a high interval compression ratio, and frequent overlapping and conflicting events. It is used to characterize the system's response continuity and conflict characteristics under high load conditions.

[0105] In this embodiment, the low-density chain propagation feature set is a feature set used to characterize the system behavior when control commands arrive at low frequencies. It is formed by combining the statistical results of continuous window intervals where the number of receptions is lower than the historical average threshold, the proportion of significantly increased command intervals, and the continuous intervals of asynchronous execution between nodes. It is used to characterize the system's response lag and discontinuity characteristics under low load conditions.

[0106] In this embodiment, three or more consecutive time windows are the minimum time span constraint for determining continuous state changes. The determination method is to use three adjacent time windows as the minimum continuous interval unit to avoid interference from single window anomalies on the overall feature determination.

[0107] In this embodiment, the first threshold multiple, the second threshold multiple, the third threshold multiple, and the fourth threshold multiple are proportional constraint parameters determined based on the statistical distribution of historical operating data. Their function is to unify the changes in reception intensity and interval of different nodes at different time scales to the same judgment standard, and to distinguish between normal fluctuations and abnormally dense or sparse states.

[0108] In this embodiment, the interval compression ratio feature is the proportion of time windows where the instruction arrival interval is significantly lower than the historical average. It is formed by statistically counting the number of windows that meet the condition of the interval being lower than the historical average multiplied by a set threshold, and expressing the correspondence with the total number of windows, which is used to characterize the trend of high-density instruction stacking.

[0109] In this embodiment, the instruction conflict count is the cumulative number of execution action conflicts caused by multiple duplicate or overriding instructions existing in the same time window at the same node. It is formed by performing window-by-window statistics on conflict-marked events in the execution record to reflect execution instability under high-density conditions.

[0110] In this embodiment, the link waiting growth interval is a set of time windows in which the forwarding waiting time increases continuously. It is formed by marking the segments in the continuous time window where the forwarding waiting time gradually increases, which is used to characterize the process of the link congestion gradually worsening.

[0111] In this embodiment, the number of brightness step transitions is the number of times the brightness change amplitude of the subsequent node changes abruptly between adjacent time windows. It is formed by counting continuous segments where the brightness change difference continuously exceeds the historical average change threshold, and is used to characterize the output discontinuity under low density conditions.

[0112] In this embodiment, the multi-node fault interval is a segment in which the brightness changes of different nodes are not kept synchronously aligned on the time axis and there is a continuous difference for more than one time window. It is formed by comparing and analyzing the time offset of brightness changes between nodes and marking it, which is used to characterize the decline in the overall coordination of the chain system.

[0113] In this embodiment, the chain propagation feature vector is a unified numerical structure formed by splicing high-density feature groups and low-density feature groups in a fixed order. Its construction method is to arrange various statistical feature values ​​in sequence, which are used as the basic expression form of reinforcement learning state input.

[0114] In this embodiment, the state input is the state representation of the chain propagation feature vector in the reinforcement learning framework. It is constructed by inputting the overall characteristics of the current chain propagation behavior of the system into the evaluation mechanism to describe the current system operating environment.

[0115] In this embodiment, the action input is the decision variable of the second light node information receiving density sequence in the reinforcement learning process. Its function is to represent the control strategy selection corresponding to different density configuration schemes, and to drive the system to execute different chain propagation methods.

[0116] In this embodiment, the feedback source is a set of evaluation indicators used to measure the quality of the current control strategy. It is composed of three types of indicators: continuous offset of brightness change, number of command conflicts, and link forwarding delay fluctuation, which are used to comprehensively reflect the control effect.

[0117] In this embodiment, the continuous offset of brightness change is the absolute accumulation of the difference between the brightness change amplitude within adjacent time windows and the average historical stable change amplitude. It is formed by successively accumulating the difference between the brightness change and the historical average for each time window, and is used to characterize the smoothness of the control output.

[0118] In this embodiment, the number of instruction conflicts is the cumulative number of conflict events caused by duplicate or overridden instructions during execution. It is formed by counting and accumulating each conflict triggering event to characterize the overload level under high-density control.

[0119] In this embodiment, the link forwarding delay fluctuation is the cumulative dispersion of the change in the time difference of instruction arrival between adjacent nodes. It is formed by statistically analyzing the change in the forwarding delay difference between each node, and is used to characterize the propagation instability under low-density conditions.

[0120] In this embodiment, the successive interactive update is a process of iteratively adjusting the control strategy based on the current state input, action input, and feedback source. It is formed by cyclically evaluating and updating the evaluation value of the control execution result corresponding to each group of second light node information receiving density sequence, so as to gradually optimize the control strategy output.

[0121] It should be noted that this step separates and models the high-density and low-density dual-feature structures in the chain propagation process, enabling the system to simultaneously identify two extreme states: instruction overload and instruction sparsity. Furthermore, it achieves comparability of different density strategies through a unified state representation method.

[0122] It should be noted that this step introduces multiple sources of reward, such as the number of conflicts and delay fluctuations, so that the evaluation mechanism not only considers the correctness of the execution result, but also the stability and continuity of the propagation process, thereby avoiding strategy bias caused by a single indicator.

[0123] It should be noted that the beneficial effect of this step is that by combining the chain propagation feature with the reinforcement learning iterative mechanism, the multi-lamp control strategy can automatically converge to the optimal control direction that balances smoothness and stability under different information reception densities, thereby improving the overall system's adaptive adjustment capability and link execution reliability.

[0124] S6. The information receiving density sequence of the second lamp node is compared step by step according to the time window value to generate an index. The index is used as the horizontal axis and the multi-lamp control reward value is used as the vertical axis to fit a curve. The information receiving density sequence of the second lamp node corresponding to the maximum multi-lamp control reward value is selected. If there are multiple sequences, the one with the smallest window change is selected as the optimal node information receiving density sequence and applied to multi-lamp control.

[0125] In this embodiment, the second light node information receiving density sequence is compared step by step according to the time window value to generate an index. A curve is fitted with the index as the horizontal axis and the multi-light control reward value as the vertical axis. The second light node information receiving density sequence corresponding to the maximum multi-light control reward value is selected. If multiple sequences exist, the one with the smallest window change is selected as the optimal node information receiving density sequence and applied to multi-light control. Specifically: The received density sequence of each second light node information is input into the sequence set as an independent sequence unit; Performing a level-wise lexicographical sort on a set of sequences includes: First, compare the received strength values ​​corresponding to the first time window in each sequence, and place the sequence with the smallest value at the front of the sequence set. When the received strength values ​​in the first time window are the same, the received strength values ​​in the corresponding second time window are compared sequentially. If the second time window is still the same, the received strength values ​​of subsequent windows are compared step by step according to the time window order. Until all time window comparisons are completed, a uniquely sorted set of the second light node information reception density sequence is obtained; The sorted set of second light node information receiving density sequences is numbered according to the arrangement order to generate a sequence index set; Each sequence index in the sequence index set is used as the x-axis value, and the single multi-light control reward value corresponding to the sequence is used as the y-axis value. Constructing sequence indexes—controlling the reward value mapping curve; Perform the following filtering process on the sequence index-control reward value mapping curve: Set the maximum reward value as the baseline value, and iterate through all sequence indices from left to right to obtain the corresponding reward values. Record the set of sequence indices that satisfy the condition that the reward value equals the baseline value; If the set has multiple sequence indices, then the following judgment is performed on the received density sequence of the second light node information corresponding to the sequence in the set: Compare the changes in received intensity values ​​for each sequence in turn for each time window, and calculate the sum of the absolute values ​​of adjacent differences in each time window; The sequence with the smallest sum of corresponding absolute values ​​is selected as the optimal second light node information receiving density sequence; If there are no multiple sequence indices that satisfy the condition that the reward value is equal to the baseline value, then the sequence with the largest corresponding reward value is directly selected as the optimal second light node information receiving density sequence. The optimal second light node information reception density sequence is applied to the grouped chain topology multi-light management and control process.

[0126] In this embodiment, the sequence set is a structured set that stores and manages all second light node information receiving density sequences in a unified manner. It is constructed by entering each second light node information receiving density sequence in the same data space according to the sequence number, which serves as a unified input object for subsequent sorting and evaluation.

[0127] In this embodiment, the layer-by-layer dictionary sorting is a sorting method that compares the received intensity values ​​of each sequence step by step according to the time window. The implementation method is to first compare the received intensity values ​​of each sequence in the first time window and sort them according to the size rule. When there are equal values ​​in the first time window, the next time window is compared in turn until a difference appears in a certain time window. This is used to establish a unique and definite priority order relationship among multiple sequences.

[0128] In this embodiment, the first time window reception intensity value is the numerical expression corresponding to the first time window in the information reception density sequence of each second light node. It is formed by the normalized expression value of the reception intensity of each node in the sequence within the initial time window, and is used as the basis for the initial sorting of the sequence.

[0129] In this embodiment, the second time window reception strength value is the reception strength value corresponding to the second time window in the sequence. It is formed by uniformly expressing the reception strength of each node in the second time window, and is used to further distinguish the sequence priority when the first window is equal.

[0130] In this embodiment, the sequential comparison of time windows is a process of comparing the received strength window by window from front to back according to the time axis. The implementation method is to compare the received strength values ​​of each time window horizontally until the position of the difference window is found, thereby determining the sequence sorting result.

[0131] In this embodiment, the sequence index set is an index structure formed by sequentially numbering the received density sequence of the second light node information after sorting. It is constructed by assigning integer numbers sequentially according to the sorting position, so that each sequence has a unique index identifier, which is used to establish a correspondence with the multi-light control reward value.

[0132] In this embodiment, the sequence index-control reward value mapping curve is a relational expression structure with the sequence index as the horizontal axis and the corresponding multi-lamp control reward value as the vertical axis. It is constructed by arranging each sequence number and its corresponding reward value in a one-to-one correspondence and connecting them in a two-dimensional coordinate system, which is used to describe the distribution trend of control effect corresponding to different receiving density sequences.

[0133] In this embodiment, the maximum reward value is the highest value among the multi-lamp control reward values ​​corresponding to all second lamp node information reception density sequences. It is determined by comparing all reward values ​​one by one to determine the maximum value, which is used as a benchmark reference for optimal sequence selection.

[0134] In this embodiment, the benchmark value is the highest standard of reward value used to screen the optimal sequence. Its function is to limit the subsequent screening to comparison only within the range of sequences that achieve the highest control effect, so as to avoid inefficient sequences from entering the final selection process.

[0135] In this embodiment, the set is a set of sequence indices whose multi-light control reward values ​​are equal to the baseline value. It is formed by comparing all sequence reward values ​​and then filtering out the indices corresponding to the same maximum value, which is used to process multiple optimal candidate cases.

[0136] In this embodiment, the variation amplitude of the received intensity value is an absolute expression of the difference in received intensity values ​​between adjacent time windows of the same sequence. It is formed by comparing the received intensity of adjacent windows one by one, taking the difference and accumulating it, which is used to measure the stability of the sequence variation.

[0137] In this embodiment, the sum of absolute values ​​is the total change obtained by accumulating the absolute values ​​of adjacent differences in each time window. It is formed by continuously accumulating the change amplitude of the entire sequence time window, which is used to characterize the overall fluctuation of the sequence.

[0138] In this embodiment, the sequence with the smallest window change is the one with the most stable change in receiving intensity under the condition of satisfying the maximum reward value. It is determined by selecting the sequence with the smallest sum of absolute values ​​among the candidate sequences, in order to ensure the continuity and stability of the control process.

[0139] It should be noted that this step involves jointly sorting and mapping the second light node information receiving density sequence with the multi-light control reward value, so that the advantages and disadvantages of different control strategies can be compared under a unified index system, thereby avoiding selection bias caused by a single time window or single node information.

[0140] It should be noted that when there are multiple optimal reward values, this step introduces a secondary screening mechanism based on the principle of minimizing window changes, so that the final selection result not only satisfies the optimal control effect, but also satisfies the stability constraints of the execution process, thereby improving the robustness of chain control.

[0141] It should be noted that the beneficial effect of this step is that by constructing an index-reward value mapping curve and combining it with a two-layer screening mechanism, the multi-lamp control system can stably select the optimal solution that combines high control effect and low fluctuation characteristics from multiple candidate density sequences, thereby improving the stability and consistency of the overall multi-lamp chain control.

[0142] Example 2: This invention also includes an intelligent multi-lamp management and control system, comprising an instruction reconstruction module, a density perturbation module, a propagation mapping module, a chain execution module, a feedback evaluation module, and a curve optimization module. The instruction reconstruction module acquires a sequence of control instruction events within a historical unit of time, rearranges them according to gateway—first node—successor node, divides the received segments according to time windows, counts the number of instruction arrivals and interval compression of each segment, and superimposes them to obtain the first lamp node information received density sequence. The density perturbation module performs several random continuous compression and stretching processes on the time window intensity of the first lamp node information received density sequence, and rearranges and maps it to chain node positions to form multiple sets of second lamp node information received density sequences. The propagation mapping module converts the node information received density sequence into a first-lamp node information received density sequence based on a segmented, time-aligned, chain-propagation mapping. The system includes a node propagating its coverage area and the subsequent node's receiving trigger window; a chain execution module applying the information receiving density sequence of each second light node to the grouped chain topology, with the gateway connecting to the first node, and the first node sequentially connecting to subsequent nodes via links, performing hierarchical forwarding and receiving control based on the coverage area and trigger window to complete the multi-light activation and deactivation; a feedback evaluation module extracting features from each multi-light control and iteratively updating the corresponding multi-light control reward value based on the Q value; and a curve optimization module comparing the information receiving density sequence of the second light node according to the time window value to generate an index, fitting a curve with the index as the horizontal axis and the multi-light control reward value as the vertical axis, selecting the second light node information receiving density sequence corresponding to the maximum multi-light control reward value, and if multiple sequences exist, selecting the one with the smallest window change as the optimal node information receiving density sequence and applying it to the multi-light control.

[0143] The present invention also includes an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform an intelligent multi-lamp management control method.

[0144] The present invention also includes a computer-readable storage medium storing a computer program that, when executed by a processor, implements an intelligent multi-lamp management and control method.

[0145] In the embodiments provided by this invention, it should be understood that the disclosed system can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0146] Furthermore, the functional modules 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. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0147] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within the invention. No appended diagram markings in the claims should be construed as limiting the scope of the claims.

[0148] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0149] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0150] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

[0151] In the embodiments provided in this disclosure, it should be understood that the disclosed systems and methods can also be implemented in other ways. The system embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0152] It should be noted that, in this disclosure, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element limited by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0153] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for intelligent multi-lamp management and control, characterized in that, Includes the following steps: The sequence of control command events within a historical unit of time is obtained, rearranged according to gateway-first node-successor node, and divided into receiving segments according to time window. The number of command arrivals and interval compression of each segment are counted and superimposed to obtain the information receiving density sequence of the first light node. The time window intensity of the first light node information reception density sequence is subjected to several random continuous compression and stretching processes, and then rearranged and mapped to the chain node positions to form multiple sets of second light node information reception density sequences. Based on the segmented preservation mapping with chain propagation time alignment, the node information reception density sequence is converted into the propagation coverage from the first node to the successor node and the reception trigger window of the successor node. The information receiving density sequence of each second light node is applied to the grouped chain topology. The gateway is connected to the first node, and the first node is connected to the subsequent nodes in sequence via links. Based on the coverage area and trigger window, hierarchical forwarding and receiving control is performed to complete the multi-light on / off operation. Feature extraction is performed on each multi-lamp control, and the corresponding multi-lamp control reward value is generated by iterative updating based on the Q value; The second light node information receiving density sequence is compared step by step according to the time window value to generate an index. A curve is fitted with the index as the horizontal axis and the multi-light control reward value as the vertical axis. The second light node information receiving density sequence corresponding to the maximum multi-light control reward value is selected. If there are multiple sequences, the one with the smallest window change is selected as the optimal node information receiving density sequence and applied to multi-light control.

2. The intelligent multi-lamp management and control method according to claim 1, characterized in that, The process involves acquiring the control command event sequence within a historical unit of time, rearranging it according to the gateway-first node-successor node sequence, dividing the received segments according to time windows, and statistically analyzing the number of command arrivals and interval compression amounts for each segment to obtain the first light node information received density sequence. Specifically: Obtain the sequence of control command events received by each light node in the grouped chain topology within a unit time, and divide the received events of each light node into a set of continuous time window segments according to a fixed-length time window. For each time window segment, the number of instruction arrivals is extracted, and the average convergence amplitude of the time interval between two adjacent instruction arrivals within each window is calculated. The average convergence amplitude is the ratio of the sum of the absolute values ​​of the differences between adjacent time intervals to the number of intervals. The number of instruction arrivals is scaled by the maximum and minimum value range, and the average convergence amplitude is shifted by the mean center, so that the two are mapped to the same numerical interval endpoint range. Within the same time window, the difference between the number of arrivals of the mapped instructions and the average convergence amplitude is statistically analyzed item by item, and the difference between each time window is accumulated and summed segment by segment to obtain the received strength value corresponding to the time window. The average of the received strength values ​​corresponding to each time window is used to form the first light node information received density for the corresponding light node. Following the chained order of gateway—first node—successor node, the first light node information reception density of each light node is arranged sequentially to form the first light node information reception density sequence.

3. The intelligent multi-lamp management and control method according to claim 2, characterized in that, The time window intensity of the first light node information reception density sequence is subjected to several random continuous compression and stretching processes, and then rearranged and mapped to the chain node positions to form multiple sets of second light node information reception density sequences, specifically as follows: For each node density value in the received density sequence of the first light node information, extract the corresponding time window intensity array in a fixed order, and divide the corresponding array into a continuous local window sub-segment with a length of three. For each local window segment, a transformation operation is performed. The transformation operation includes shifting the intensity difference between adjacent windows in the segment sequentially backward by one time step to generate a compressed transformation segment, and inserting an equal number of zero compensation terms into the intensity difference between adjacent windows in the segment sequentially to generate a stretched transformation segment. The compression transformation segments and the stretch transformation segments are alternately combined according to the original time window order to form a reconstruction sequence of several time windows corresponding to the nodes; For each time window reconstruction sequence, the topological position index of the gateway-first node-successor node is randomly rearranged so that the intensity of each time window in the same sequence corresponds one-to-one with the corresponding node number, forming a node mapping sequence. Based on the time window strength of the first node in the node mapping sequence, sequential delay alignment is performed on the time window strength of its successor nodes to ensure that the time window strength of all nodes is expressed with the same time step length. The aligned node mapping sequence is output as a set of second lamp node information receiving density sequences, and the alignment is repeated for all reconstructed sequences to form several sets of second lamp node information receiving density sequences.

4. The intelligent multi-lamp management and control method according to claim 3, characterized in that, The segmented preservation mapping based on chain-propagation time alignment converts the node information reception density sequence into the propagation coverage area from the first node to the subsequent nodes and the reception trigger window of the subsequent nodes, specifically as follows: The starting time marker corresponding to the continuous expression sequence of the received intensity per unit time of each light node is set as the relative time when the node accesses the gateway, and the time when the gateway issues the first control command is used as the zero point reference of the global time axis for chain propagation. Starting from the zero point of the global time axis, the time offset of each time window in the continuous expression sequence of the received intensity per unit time of each lamp node is calibrated. The time offset is determined by the time difference between the time when the corresponding node accesses the gateway and the zero point of the global time axis. Subtract the corresponding time offset from the start and end boundaries of each time window in the time window sequence of each light node, so that the time window sequence of all nodes is converted to a unified global time axis scale system. Under a unified global timeline scale system, the time windows of different light nodes are rearranged segment by segment according to the time window start-point alignment rule, so that the time windows within the same time scale range have a consistent start and end boundary expression on the global timeline. On the continuous expression sequence of received intensity per unit time after global time axis alignment, based on the principle of keeping the original time window boundaries unchanged, the continuous expression sequence of each lamp node is divided into segments according to the start and end boundaries of the corresponding time window, forming a segmented sequence that corresponds one-to-one with the original time window. For each hold segment, the segment duration and segment intensity peak are extracted. The segment duration is determined by the duration span of the corresponding time window on the global time axis, and the segment intensity peak is determined by the maximum value in the continuous expression sequence of received intensity per unit time within the corresponding segment. The segment duration is used as a measure of propagation coverage capability, and the segment intensity peak is used as a benchmark for propagation intensity constraints. The propagation coverage capability representation is passed level by level in a chain topology order from gateway to first node to successor node, and at each successor node, the propagation coverage capability representation is updated by decreasing according to the link propagation delay constraint between adjacent nodes, generating a set of continuous coverage nodes on the corresponding chain topology path. Within the holding segment corresponding to each successor node, a symmetrical time expansion interval is constructed on the global time axis, with the time position corresponding to the peak intensity of the segment as the reference point, and the interval is used as the receiving trigger time interval of the corresponding node. By binding and matching the continuous set of covered nodes with the corresponding receiving trigger time intervals of each successor node on the global time axis, a segmented and persistent mapping structure based on chain propagation timing alignment is formed.

5. The intelligent multi-lamp management and control method according to claim 4, characterized in that, The method involves applying the information reception density sequence of each second light node to a packet chain topology. The gateway connects to the first node, and the first node connects to subsequent nodes sequentially via links. Based on the coverage area and trigger window, hierarchical forwarding and reception control is performed to complete the multi-light activation and deactivation. Specifically: For each second light node information receiving density sequence, a group-level control instruction packet is first generated on the gateway side, and the corresponding propagation coverage range parameter and subsequent node receiving trigger window parameter are written into the instruction packet, while the group identifier and sequence identifier are attached. The gateway sends the instruction packet to the first node of the corresponding light group. After receiving it, the first node performs chain mapping decomposition on the propagation coverage parameters, generates the local coverage segment and the next-hop forwarding segment corresponding to the first node, and splits the received trigger window parameters into time intervals to generate its own execution window and the next-level forwarding window. The first node performs control actions within the trigger time interval according to its own execution window, and initiates a chain-like forwarding process within the coverage area, synchronously sending control commands to the next successor node; After each successor node receives the instruction, it confirms whether it is within the effective coverage range based on its corresponding coverage segment parameters, and determines whether to enter the execution state window based on the received trigger window parameters. When the conditions are met, it performs the corresponding opening / closing or dimming action, and synchronously generates the next hop coverage segment and trigger window. Each successor node repeats the above-mentioned coverage segment inheritance and trigger window inheritance process level by level along the chain topology, so that the propagation coverage range is passed along the link level by level and decays level by level, while the trigger window remains continuously aligned and propagates on the time axis. When the coverage area propagates to the end node or the coverage segment decays to the preset termination condition, forwarding stops, completing the multi-lamp on / off control of the chain control execution process corresponding to the information reception density sequence of a single second lamp node.

6. The intelligent multi-lamp management and control method according to claim 5, characterized in that, The process of extracting features from each multi-light control and iteratively updating the corresponding multi-light control reward value based on the Q value is as follows: Based on the chain control execution process corresponding to the information reception density sequence of each second light node, the sequence of instruction reception times, the sequence of repeated reception times, and the sequence of adjacent instruction arrival time intervals of each node are extracted in the propagation order of gateway, first node and subsequent nodes. The sequence of instruction coverage overlap times and forwarding waiting time of each node in the same time window is also recorded to form node-level propagation execution characteristics. The node-level propagation execution features are merged and spliced ​​in a chain-like topological order to form a chain-like propagation state representation corresponding to the information reception density sequence of the second light node; In the chain propagation state representation, a feature decomposition set is constructed to characterize the differences in receiver density, including a high-density chain propagation feature set and a low-density chain propagation feature set, specifically: For high-density chain propagation feature groups, based on the sequence of instruction reception counts within a unit time window, the length of the continuous interval corresponding to the first threshold multiple of the reception count of each window in three or more consecutive time windows is extracted as the high-density continuous reception interval length, which is used to characterize the execution feature of continuous brightness adjustment and fine response when the reception density is high. Based on the instruction arrival interval sequence within the same time window, the percentage of times the interval value is lower than the second threshold multiple of the historical average interval of the node is statistically analyzed to form the interval compression ratio feature, which is used to characterize the state of rapid and continuous instruction superposition under high density conditions. Based on the overlap sequence, the number of conflicts occurs when the overlap number exceeds the maximum number of conflict-free overlaps in the history of a node within a single time window, forming an instruction conflict frequency feature to characterize the conflicts caused by the superposition of new and old instructions due to high density. Based on the forwarding wait time sequence, the number of intervals in which the wait time continuously increases within three or more consecutive time windows is extracted to form the link wait growth interval feature, which is used to characterize the information overload and forwarding blocking status of intermediate nodes. For low-density chain propagation feature groups, based on the sequence of instruction reception counts within a unit time window, the length of the continuous interval corresponding to the third threshold multiple when the number of receptions in each of three or more consecutive time windows is lower than the average number of receptions in the historical windows of the node is extracted, and used as the length of the low-density continuous sparse interval to characterize the state of reduced node load but insufficient response. Based on the instruction arrival interval sequence, the percentage of times the interval value is higher than 1.8 times the historical average interval of the corresponding node is counted to form a high interval percentage feature, which is used to characterize the sparse instruction propagation and response delay state; based on the brightness change sequence of the successor node, the number of consecutive times when the difference between the brightness change amplitude of two adjacent time windows continuously exceeds the fourth threshold of the historical average change amplitude is counted to form a brightness step jump number feature, which is used to characterize the brightness change as a segmented jump state. Based on the time alignment difference between brightness change sequences of different nodes, the number of asynchronous duration intervals exceeding one time window length is counted to form multi-node tomographic interval features, which are used to characterize the discontinuous brightness state of multi-lamp regions. The high-density chain propagation feature group and the low-density chain propagation feature group are combined in a fixed order to form a chain propagation feature vector corresponding to the information receiving density sequence of each second lamp node. The second light node information receiving density sequence is used as the action input, the chain propagation feature vector is used as the state input, and the continuous offset of brightness change of each node, the number of command conflicts and the link forwarding delay fluctuation during the chain control execution process are used as the feedback sources. The continuous offset of brightness change is the absolute accumulation of the difference between the brightness change amplitude of adjacent time windows and the average of the historical stable change amplitude, which is used to characterize the control smoothness. The number of instruction conflicts is the count of execution conflicts caused by repeated instruction overwriting within the same time window, used to characterize high-density overload penalties; Link forwarding delay fluctuation is the cumulative variance of the instruction arrival time difference between adjacent nodes, used to characterize the low-density propagation fault penalty; Based on the state input, action input, and reward source, the chain control execution process corresponding to the information receiving density sequence of each second light node is updated sequentially to generate the corresponding single multi-light control reward value.

7. The intelligent multi-lamp management and control method according to claim 6, characterized in that, The process involves comparing the received density sequence of the second light node information step by step according to the time window value to generate an index. A curve is fitted with the index as the horizontal axis and the multi-light control reward value as the vertical axis. The received density sequence of the second light node corresponding to the maximum multi-light control reward value is selected. If multiple sequences exist, the one with the smallest window change is selected as the optimal node information received density sequence and applied to multi-light control. Specifically: The received density sequence of each second light node information is input into the sequence set as an independent sequence unit; Performing a level-wise lexicographical sort on a set of sequences includes: First, compare the received strength values ​​corresponding to the first time window in each sequence, and place the sequence with the smallest value at the front of the sequence set. When the received strength values ​​in the first time window are the same, the received strength values ​​in the corresponding second time window are compared sequentially. If the second time window is still the same, the received strength values ​​of subsequent windows are compared step by step according to the time window order. Until all time window comparisons are completed, a uniquely sorted set of the second light node information reception density sequence is obtained; The sorted set of second light node information receiving density sequences is numbered according to the arrangement order to generate a sequence index set; Each sequence index in the sequence index set is used as the x-axis value, and the single multi-light control reward value corresponding to the sequence is used as the y-axis value. Constructing sequence indexes—controlling the reward value mapping curve; Perform the following filtering process on the sequence index-control reward value mapping curve: Set the maximum reward value as the baseline value, and iterate through all sequence indices from left to right to obtain the corresponding reward values. Record the set of sequence indices that satisfy the condition that the reward value equals the baseline value; If the set has multiple sequence indices, then the following judgment is performed on the received density sequence of the second light node information corresponding to the sequence in the set: Compare the changes in received intensity values ​​for each sequence in turn for each time window, and calculate the sum of the absolute values ​​of adjacent differences in each time window; The sequence with the smallest sum of corresponding absolute values ​​is selected as the optimal second light node information receiving density sequence; If there are no multiple sequence indices that satisfy the condition that the reward value is equal to the baseline value, then the sequence with the largest corresponding reward value is directly selected as the optimal second light node information receiving density sequence. The optimal second light node information reception density sequence is applied to the grouped chain topology multi-light management and control process.

8. A system using the intelligent multi-lamp management and control method as described in any one of claims 1-7, characterized in that, It includes an instruction reconfiguration module, a density perturbation module, a propagation mapping module, a chain execution module, a feedback evaluation module, and a curve optimization module; The instruction reconstruction module is used to obtain the sequence of control instruction events within a historical unit of time, rearrange them according to the gateway-first node-successor node, divide the received segments according to the time window, count the number of instruction arrivals and the interval compression amount of each segment and superimpose them to obtain the first light node information received density sequence. The density perturbation module is used to perform several random and continuous compression and stretching processes on the time window intensity of the first lamp node information receiving density sequence, and rearrange and map it to the chain node position to form multiple sets of second lamp node information receiving density sequences. The propagation mapping module is used for segmented preservation mapping based on chain propagation time alignment, which converts the node information reception density sequence into the propagation coverage from the first node to the subsequent nodes and the reception trigger window of the subsequent nodes. The chain execution module is used to apply the information reception density sequence of each second light node to the group chain topology. The gateway is connected to the first node, and the first node is connected to the subsequent nodes in sequence through the link. According to the coverage and trigger window, the module performs hierarchical forwarding and reception control to complete the multi-light opening and closing. The feedback evaluation module is used to extract features from each multi-lamp control and generate the corresponding multi-lamp control reward value based on the Q value through iterative updates. The curve optimization module is used to compare the second light node information reception density sequence step by step according to the time window value to generate an index. The index is used as the horizontal axis and the multi-light control reward value is used as the vertical axis to fit a curve. The second light node information reception density sequence corresponding to the maximum multi-light control reward value is selected. If there are multiple sequences, the one with the smallest window change is selected as the optimal node information reception density sequence and applied to multi-light control.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, which is then executed by the at least one processor to enable the at least one processor to perform the intelligent multi-lamp management and control method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent multi-lamp management and control method as described in any one of claims 1 to 7.

Images