A multi-node lighting device cooperative control method and system

Abstract

This invention discloses a collaborative control method and system for multi-node lighting equipment, relating to the field of collaborative control technology. The invention collects multi-dimensional state data from the target, execution, and adjacent nodes, calculates consistency correlation coefficients and state deviations, constructs a virtual impedance, and couples it with a pseudo-random sequence to form a time-varying impedance. The time-varying impedance is used to adjust control commands, generating actual driving voltages with physical watermarks to drive the execution nodes. The physical watermark response is then analyzed using a polarity cross-correlation function to distinguish between real physical nodes and virtual deception nodes. The correlation coefficient for the next round of control is updated based on a trust penalty coefficient. This invention effectively prevents equipment overload and energy waste caused by abnormal commands, improving the reliability and anti-interference capability of the collaborative control network.

Description

Technical Field

[0001] This invention relates to the field of collaborative control technology, and in particular to a collaborative control method and system for multi-node lighting equipment. Background Technology

[0002] With the widespread application of intelligent lighting systems in smart parks, smart buildings, and other scenarios, the collaborative control of multi-node lighting devices has become a key technology for improving lighting quality, optimizing energy utilization, and ensuring the safe and stable operation of the system. This type of technology, through information interaction and collaboration between multiple lighting devices, can dynamically adjust according to environmental changes, pedestrian activity, and user needs, thereby reducing energy consumption while ensuring lighting comfort. However, in actual deployments, lighting nodes are usually widely distributed, and the operating environment is complex and variable. The states between nodes may be inconsistent due to sensor errors, communication delays, or equipment aging. In addition, in open communication networks, the system may also face the risk of malicious nodes sending false commands or injecting abnormal data. These factors can affect the accuracy and reliability of collaborative control, and may even lead to equipment damage or energy waste. In the prior art, Chinese invention patent application number CN120111750A discloses a lighting control method, device, equipment, and storage medium. This scheme generates energy efficiency allocation strategies and preliminary control strategies based on environmental perception data through metabolic decision branches and neural decision branches in a collaborative decision engine. It then uses a digital twin verification platform to perform virtual simulation and parameter optimization of the strategies, and adjusts lighting parameters by combining a physiological response prediction model to achieve efficient collaboration between metabolic energy efficiency allocation and neural control decisions. However, this scheme mainly focuses on single-engine optimization and physiological response prediction based on metabolic-neural coupling. In the process of multi-node collaborative control, it does not adequately consider real-time consistency verification of state data between nodes, dynamic adjustment of instructions based on state deviation, and physical layer verification against malicious attacks. In particular, when facing abnormal situations such as node state distortion, instruction tampering, or injection of false data, this scheme lacks an effective real-time detection and adaptive suppression mechanism, which may lead to control strategy distortion, energy allocation errors, and even affect the safe and stable operation of the lighting system. These defects limit the prior art in multi-node collaborative lighting control scenarios that require high reliability and high security. Summary of the Invention

[0003] The technical problem solved by this invention is that in the process of multi-node collaborative control, the existing technology does not adequately consider the real-time consistency verification of the state data between nodes and the physical layer verification against malicious attacks. In particular, when facing abnormal situations such as node state distortion, instruction tampering or false data injection, there is a lack of effective real-time detection and adaptive suppression mechanisms, which may lead to distortion of control strategy, incorrect energy allocation, and even affect the safe and stable operation of the lighting system.

[0004] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for collaborative control of multi-node lighting devices, comprising the following steps: Step S1: Collect control commands from the target node and collect multi-dimensional status data of the target node, execution node, and adjacent nodes; Step S2: Calculate the correlation coefficient between the multi-dimensional state data of the target node, the execution node and the neighboring nodes. Based on the multi-dimensional state data of the neighboring nodes, filter the effective dimensions in the multi-dimensional state data. Substitute the effective dimension state data of the target node into the consistency differential equation weighted by the correlation coefficient and iterate until convergence. Calculate the state deviation. Step S3: Construct virtual resistance parameters based on the amplitude of the state deviation, construct virtual inductance parameters based on the rate of change of the state deviation over time, the virtual resistance parameters and virtual inductance parameters constitute virtual impedance, construct a pseudo-random sequence, and couple the pseudo-random sequence with the virtual impedance to construct a time-varying impedance; Step S4: Adjust the control command of the target node using the time-varying impedance, generate an actual drive command with a physical watermark to drive the execution node, and update the correlation coefficient according to the target node's response to the physical watermark.

[0005] Preferably, in step S1, the target node is the source lighting device node that initiates the control command; The execution node is a lighting device node that receives the control command and is responsible for performing physical lighting actions; The adjacent nodes are other lighting device nodes that have a direct physical communication connection with the execution node; The control command includes the drive voltage and duration; The multi-dimensional state data includes voltage-dimensional state data, current-dimensional state data, ambient light-dimensional state data, moving object detection-dimensional state data, and moving object speed-dimensional state data. The voltage dimension state data is a sequence of instantaneous working voltage values ​​obtained by sorting the instantaneous working voltage values ​​by time within a time window; The current dimension state data is a sequence of instantaneous operating current values ​​obtained by sorting the operating current values ​​within a time window by time. The ambient light intensity dimension status data is an ambient light intensity sequence obtained by sorting the ambient light intensity by time within a time window; The moving object detection dimension state data is a sequence of moving object detection states obtained by sorting the moving object detection states by time within a time window. The state data for the velocity dimension of the moving object is a velocity sequence obtained by sorting the detected velocities of the moving object within a time window by time.

[0006] Preferably, in step S2, the process of calculating the consistency of the multi-dimensional state data of the target node, the execution node, and the adjacent nodes specifically includes: Extract multi-dimensional state data from the target node, execution node, and adjacent nodes respectively; The state data of the target node in each dimension are the state data of each dimension within a preset time window before the control command is generated; The status data of each dimension of the execution node are the status data of each dimension within a preset time window before receiving the control command from the target node; The state data of each dimension of the adjacent node is the state data of each dimension within a preset time window before the executing node receives the control command from the target node. Calculate the correlation coefficient between the state data of adjacent nodes and the execution node in each dimension; If the correlation coefficient between the state data of adjacent nodes and execution nodes is greater than the preset first correlation threshold, the corresponding dimension is determined to be a valid dimension, and the state data of the valid dimensions of the target node and execution node are constructed into a valid state set.

[0007] Preferably, in step S2, the process of calculating the consistency of the state data of the target node, the execution node, and the adjacent nodes further includes: Within the set of valid states, calculate the correlation coefficients between the multi-dimensional state data of the target node and the multi-dimensional state data of the execution node, respectively. The target node and its adjacent nodes are used as connection nodes for the execution node. The connection weights between the execution node and the connection nodes are constructed, and the initial value of each connection weight is 1. Using the correlation coefficients of execution nodes and connected nodes under the effective dimension, the connection weights are weighted and corrected to obtain the weighted connection weights. The mathematical expression for the weighted connection weights is as follows: ; in, For the corrected connection weights, The initial values ​​for the connection weights are... For execution node With the target node and neighboring nodes The correlation coefficient; The multi-dimensional state data of the target node, execution node, and adjacent nodes at a certain moment are vertically concatenated into a column vector in the effective dimension to obtain the instantaneous state vector at that moment. The instantaneous state vector is then substituted into a first-order consistency differential equation for consistency verification. The mathematical expression of the consistency differential equation is as follows: ; in, The change in the state data of the execution node. It is the set of neighboring nodes. Let be the column vector of the neighboring nodes at time t. This is the column vector of the execution node at time t; Iterative calculations are performed until convergence, yielding consistent convergence state data. The state deviation is obtained by calculating the Euclidean distance between the state data of the target node and the consistent converged state data.

[0008] Preferably, in step S2, the process of constructing the virtual impedance specifically includes: A virtual impedance is constructed based on the state deviation, and the virtual impedance includes virtual resistance parameters and virtual inductance parameters. The mathematical expression for the virtual resistance parameter is: ; in, For virtual resistance, Based on the fundamental resistivity, This is the resistance gain factor. This refers to the degree of state deviation. The virtual inductance parameters : ; in, Based on the fundamental inductance coefficient, The rate of change of the deviation from the state over time. This represents the degree of state deviation.

[0009] Preferably, in step S3, the process of adjusting the control command of the target node specifically includes adjusting the driving voltage and duration, and obtaining the actual driving voltage and actual duration. A binary pseudo-random M-sequence within a preset time window is generated using a linear feedback shift register, and the binary pseudo-random M-sequence is used as a modulation factor and coupled with a virtual impedance to form a time-varying impedance. The actual driving voltage is obtained by calculating the virtual voltage drop generated by the time-varying impedance in the execution node circuit. The mathematical expression for the actual driving voltage is as follows: ; in, This is the actual driving voltage. The drive voltage value in the control command sent to the target node. For virtual resistance parameters, For virtual inductance parameters, For the execution node at time Instantaneous operating current, This is the modulation depth coefficient (in this embodiment, the value range is 0.05~0.1). The mathematical expression for the actual duration is: ; in, This refers to the actual execution window duration. The time sensitivity coefficient (in this embodiment, the value is 10). To control the duration in the command, This represents the degree of state deviation.

[0010] Preferably, in step S3, the process of updating the correlation coefficient based on the target node's response to the physical watermark specifically includes: The working voltage instantaneous value sequence of the target node within a preset time window is extracted and constructed into a working voltage instantaneous value vector. The binary pseudo-random M sequence and the working voltage instantaneous value vector are subjected to binarization polarity processing to obtain the local polarity reference sequence and the target polarity measurement sequence. The polarity cross-correlation function between the local polarity reference sequence and the target polarity measurement sequence is calculated using logical operations. The mathematical expression of the polarity cross-correlation function is as follows: ; in, The value of the polar cross-correlation function. This represents the total number of sampling points within the time window. For sliding time delay, Let k be the value of the binary pseudo-random M-sequence at time k. Let be the voltage at the target node at time k+τ. It is a symbolic function.

[0011] Preferably, in step S3, the process of updating the correlation coefficient based on the target node's response to the physical watermark further includes: If the peak value of the polar cross-correlation function is greater than the preset second correlation threshold, the target node is determined to be a physical connection node; If the peak value of the polar cross-correlation function is less than or equal to the second correlation threshold, the target node is determined to be a virtual deception node.

[0012] Preferably, in step S3, the process of updating the correlation coefficient based on the target node's response to the physical watermark further includes: Different preset trust penalty coefficients are assigned to physical connection nodes and virtual deception nodes respectively, and the relevant coefficients are updated in the next round of collaborative control cycle; The updated mathematical expression for the correlation coefficient is as follows: ; in, The original Pearson coefficients are obtained from the multi-dimensional state data collected during the next round of coordinated control. The trust penalty coefficient, This is the updated correlation coefficient.

[0013] A multi-node lighting equipment collaborative control system includes a data acquisition module, a status characteristic module, a virtual impedance module, and a control module; The acquisition module is used to acquire control commands from the target node and multi-dimensional status data of the target node, the execution node, and adjacent nodes. The state feature module calculates the consistency of multi-dimensional state data of the target node, execution node, and adjacent nodes as a correlation coefficient, and calculates the state deviation. The virtual impedance module calculates virtual impedance based on the state deviation, constructs a pseudo-random sequence, and couples the pseudo-random sequence with the virtual impedance to construct a time-varying impedance. The control module uses the time-varying impedance to adjust the control commands of the target node, generates actual drive commands with physical watermarks to drive the execution node, and updates the correlation coefficients according to the target node's response to the physical watermarks.

[0014] The beneficial effects of this invention are as follows: This invention utilizes a consensus differential equation to calculate the multi-dimensional state deviation between the target node, the execution node, and adjacent nodes, and dynamically maps this deviation to a second-order virtual impedance parameter. This mechanism, which transforms data deviation at the information layer into impedance constraints at the physical layer, effectively prevents equipment overload or energy waste caused by erroneous strategies. By modulating a uniquely identifiable physical watermark in the actual driving voltage and using a polarity cross-correlation function to calculate the waveform similarity between the local reference sequence and the feedback voltage sequence of the target node, it is possible to strictly distinguish between real physical connection nodes and virtual deception nodes that only exist at the logic layer from the physical signal level. Based on the response of the physical watermark, this invention uses a trust penalty coefficient to nonlinearly penalize the association weight of the deception node and updates the correlation coefficient in the next round of collaborative control cycle, which can ensure the high reliability and anti-interference capability of the multi-node lighting equipment collaborative control network in long-term operation. Attached Figure Description

[0015] Figure 1 The flowchart illustrates the steps of a multi-node lighting device collaborative control method according to an embodiment of the present invention. Detailed Implementation

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0017] Example 1, referring to Figure 1 A method for collaborative control of multi-node lighting devices is provided, comprising the following steps: Step S1: Collect control commands from the target node and collect multi-dimensional status data of the target node, execution node, and adjacent nodes; Step S2: Calculate the correlation coefficient between the multi-dimensional state data of the target node, the execution node and the neighboring nodes. Based on the multi-dimensional state data of the neighboring nodes, filter the effective dimensions in the multi-dimensional state data. Substitute the effective dimension state data of the target node into the consistency differential equation weighted by the correlation coefficient and iterate until convergence. Calculate the state deviation. Step S3: Construct virtual resistance parameters based on the magnitude of the state deviation, construct virtual inductance parameters based on the rate of change of the state deviation over time, and construct virtual impedance based on the virtual resistance parameters and virtual inductance parameters. Construct a pseudo-random sequence, and couple the pseudo-random sequence with the virtual impedance to construct a time-varying impedance. Step S4: Adjust the control command of the target node using time-varying impedance, generate actual drive command with physical watermark to drive the execution node, and update the correlation coefficient according to the target node's response to the physical watermark.

[0018] This invention quantifies the degree of abnormality in node behavior based on the state deviation calculation of multi-dimensional data. It transforms this deviation in the information layer into time-varying impedance in the physical layer, which not only enables adaptive and flexible adjustment of abnormal control commands, but also generates a physical watermark by coupling a pseudo-random sequence in the impedance. This allows the system to actively detect and verify the physical authenticity of the target node while performing the lighting task, effectively defending against deceptive attacks such as false data injection and non-physical connection nodes.

[0019] In step S1, the target node is the source lighting device node that initiates the control command; The execution node is the lighting equipment node that receives control commands and is responsible for executing physical lighting actions; Adjacent nodes are other lighting device nodes that have a direct physical communication connection with the execution node; The control commands include the drive voltage (which controls the brightness of the lighting equipment) and the duration; Multi-dimensional state data includes voltage-dimensional state data, current-dimensional state data, ambient light-dimensional state data, moving object detection-dimensional state data, and moving object velocity-dimensional state data; The voltage dimension status data is a sequence of instantaneous operating voltage values ​​sorted by time within a time window; The current dimension status data is a sequence of instantaneous operating current values ​​obtained by sorting the operating current values ​​within the time window by time. The ambient light intensity dimension status data is an ambient light intensity sequence obtained by sorting the ambient light intensity by time within a time window. The moving object detection dimension state data is a sequence of moving object detection states ordered by time within a time window. The state data for the velocity dimension of moving objects is a velocity sequence obtained by sorting the detected velocities of moving objects within a time window by time.

[0020] In one specific embodiment of the present invention, the lighting system adopts the ZigBeeMesh self-organizing network communication protocol. Each lighting device has a pre-stored neighbor list, which records the IDs of directly connected nodes with wireless signal strength better than -85dBm. The definition of target node, execution node and neighboring node is based on the generation of cooperative events. The radar of the target node first detects the moving object and sends a control command (cooperative request) to the execution node. The control command (cooperative request) comes from the lighting strategy preset in the target node. The lighting strategy refers to the condition trigger and control parameter mapping table stored in the target node's microprocessor. The condition trigger is when the target node's radar or camera detects a moving object. The target node generates its own control parameters and cooperative control instructions for the execution node according to the table lookup method. The cooperative control instructions include the driving voltage and duration. The execution node receives control commands and performs calculations of state deviation, time-varying impedance, and actual control commands. Adjacent nodes are used as reference adjacent nodes for the execution node to verify the target node. The speed and position information of the moving object comes from the radar and camera of the lighting equipment. The position information is relative position information, including the direction and distance relative to the target node, such as the moving object being 50 meters due north of the target node. The moving object detection status is the status of the moving objects detected by the current node, including whether there are moving objects or not. The code for detecting a moving object is 1, and the code for detecting no moving object is 0. All multi-dimensional state data still need to be normalized.

[0021] Step S2, the process of calculating the consistency of multi-dimensional state data of the target node, the execution node, and neighboring nodes specifically includes: Extract multi-dimensional state data from the target node, execution node, and adjacent nodes respectively; The status data of the target node in each dimension are the status data of each dimension within a preset time window before the control command is generated; The status data of each dimension of the execution node are the status data of each dimension within a preset time window before receiving the control command from the target node; The status data of each dimension of the adjacent nodes are the status data of each dimension within a preset time window before the executing node receives the control command from the target node; Calculate the correlation coefficient between the state data of adjacent nodes and the execution node in each dimension; If the correlation coefficient between the state data of adjacent nodes and execution nodes is greater than the preset first correlation threshold, the corresponding dimension is determined to be a valid dimension, and the state data of the valid dimensions of the target node and execution node are constructed into a valid state set.

[0022] In a specific embodiment of the present invention, the length of the preset time window is 100ms. For the target node, the state data of each dimension of the target node within the preset time window before the generation of control instructions are extracted, including: working voltage instantaneous value sequence, working voltage instantaneous value sequence, ambient light intensity sequence, moving object detection state sequence and speed sequence. Extract the status data of the execution node and its adjacent nodes in various dimensions. The time window is a preset time window before receiving the control command from the target node. Calculate the Pearson correlation coefficients between the instantaneous working voltage value sequence of the execution node and its adjacent nodes, the instantaneous working voltage value sequence, the ambient light intensity sequence, the moving object detection state sequence, and the velocity sequence, respectively; Dimensions with a Pearson correlation coefficient greater than a preset first correlation threshold (0.8 in this embodiment) are defined as valid dimensions; The state data (sequences) under the effective dimensions of the execution node and the target node are constructed into a set of effective dimensions.

[0023] This invention can automatically filter out effective dimensions with high spatial consistency by calculating the Pearson correlation coefficient of data between adjacent nodes and execution nodes within a preset time window. This step is equivalent to adding a data cleaning process before entering the core algorithm, ensuring that the subsequent consistency calculation is based on reliable and representative environmental characteristics, and avoiding misleading the judgment of the entire cooperative control strategy due to the drift or damage of individual sensors.

[0024] In step S2, the process of calculating the consistency of the state data of the target node, the execution node, and adjacent nodes also includes: Within the valid state set, calculate the correlation coefficient (Pearson correlation coefficient) between the multi-dimensional state data of the target node and the multi-dimensional state data of the execution node. The target node and its adjacent nodes are used as connection nodes for the execution node. The connection weights between the execution node and the connection nodes are constructed, and the initial value of each connection weight is 1. Using the correlation coefficients of execution nodes and connection nodes under the effective dimensions, the connection weights are weighted and adjusted to obtain the weighted connection weights. The mathematical expression for the weighted connection weights is: ; in, For the corrected connection weights, The initial values ​​for the connection weights are... For execution node With the target node and neighboring nodes The correlation coefficient; The multi-dimensional state data of the target node, execution node, and adjacent nodes at a certain moment are vertically concatenated into a column vector in the effective dimension to obtain the instantaneous state vector at that moment. The instantaneous state vector is then substituted into a first-order consistency differential equation for consistency verification. The mathematical expression of the consistency differential equation is as follows: ; in, The change in the state data of the execution node. It is the set of neighboring nodes. Let be the column vector of the neighboring nodes at time t. This is the column vector of the execution node at time t; Iterative calculations are performed until convergence, yielding consistent convergence state data. The state deviation is obtained by calculating the Euclidean distance between the state data of the target node and the consistent converged state data.

[0025] In one specific embodiment of the present invention, the change in state data of each dimension of the execution node is calculated using a consistent differential equation (adjusting the gradient). The state data of each dimension of the current execution node is updated by adding the change in state data of each dimension to the current state data of the execution node. The updated state data of each dimension is then substituted into the consistent differential equation for further calculation and updating until... Less than the preset convergence threshold (10 in this embodiment) -4 Once the consensus convergence is achieved, the state data of each dimension of the execution node calculated at this time is used as the consensus convergence state data.

[0026] This invention dynamically adjusts connection weights using correlation coefficients, giving highly correlated nodes higher trust levels. This makes the calculated consensus convergence state more representative of the real local environment level of the network. The state deviation calculated using this convergence state not only reflects the difference between the target node and the execution node, but also the difference between the target node and the consensus of the entire local network. This provides a precise mathematical basis for subsequent virtual impedance adjustment and ensures the objectivity of abnormal behavior judgment.

[0027] In step S2, the process of constructing the virtual impedance specifically includes: A virtual impedance is constructed based on the state deviation, and the virtual impedance includes virtual resistance parameters and virtual inductance parameters. The mathematical expression for the virtual resistance parameter is: ; in, For virtual resistance, The basic resistivity is 5.0Ω in this embodiment. This is the resistance gain factor (0.88 in this embodiment). This refers to the degree of state deviation. The virtual inductance parameters : ; in, The basic inductance coefficient (2.0H in this embodiment) The rate of change of the deviation from the state over time. This represents the degree of state deviation.

[0028] In one specific embodiment of the present invention, the virtual resistance parameter is exponentially positively correlated with the instruction deviation, and is used to attenuate the amplitude of the abnormal instruction in steady state; the virtual inductance parameter is positively correlated with the rate of change of the instruction deviation, and is used to suppress the sudden change rate of the abnormal instruction in transient state. The lower the state deviation, the more closely the environmental feature data provided by the target node matches the data of the neighboring nodes used as references. In this case, the exponential term... When the value is close to 1, the control commands of the target node are only slightly attenuated and can be responded to by the execution node almost in full. The greater the deviation in state, the more serious the distortion or falsification of the environmental characteristic data of the target node is. Due to the amplification characteristics of the exponential function, the virtual resistance will increase rapidly and nonlinearly. The virtual resistance forms a significant voltage drop in the droop control calculation, forcing the actual driving command to be much lower than the original control command, thereby forcibly attenuating or blocking the suspicious high brightness request at the physical level. Differentiating the time series of state deviations yields the rate of change of deviation. And substitute the mathematical expression of the virtual inductance parameters into it to process the time-domain dynamic characteristics of the control command; If the data deviation of the target node is generated slowly (such as a slow change caused by sensor drift) and the rate of change approaches 0, then the virtual inductance approaches 0. At this time, the control command only relies on the virtual resistance for amplitude correction and does not introduce phase lag. If the data deviation of the target node changes drastically in a very short time (such as a pulse attack command sent by a hacker at the moment of access), the rate of change will reach a maximum value, causing the virtual inductance to increase instantaneously. Utilizing the physical characteristic of inductance that allows DC to pass while blocking AC, the huge virtual inductance will smooth out the high-frequency sudden change components in the control command, preventing the output power of the execution node from experiencing destructive oscillations or instantaneous overshoot.

[0029] This claim constructs a virtual impedance comprising virtual resistance and virtual inductance, achieving nonlinear physical suppression of abnormal commands. The virtual resistance increases exponentially with the state deviation, significantly attenuating abnormal voltage requests with high deviations in steady state to prevent overload. The virtual inductance responds to the rate of change of deviation, effectively smoothing and suppressing sudden pulse attacks or drastic command jumps by utilizing the characteristics of inductance that pass DC and block AC. This mechanism transforms data security issues into power electronic control issues, protecting the safety of lighting hardware without disconnecting the connection.

[0030] In step S3, the process of adjusting the control command of the target node specifically includes adjusting the driving voltage and duration, and obtaining the actual driving voltage and actual duration. Using a linear feedback shift register, a binary pseudo-random M-sequence within a preset time window is generated, and the binary pseudo-random M-sequence is used as a modulation factor and coupled with a virtual impedance to form a time-varying impedance; The actual driving voltage is obtained by calculating the virtual voltage drop generated by the time-varying impedance in the execution node circuit. The mathematical expression for the actual driving voltage is: ; in, This is the actual driving voltage. The drive voltage value in the control command sent to the target node. For virtual resistance parameters, For virtual inductance parameters, For the execution node at time Instantaneous operating current, This is the modulation depth coefficient (in this embodiment, the value range is 0.05~0.1). The mathematical expression for the actual duration is: ; in, This refers to the actual execution window duration. The time sensitivity coefficient (in this embodiment, the value is 10). To control the duration in the command, This represents the degree of state deviation.

[0031] In one specific embodiment of the present invention, a binary pseudo-random M-sequence is generated using a linear feedback shift register and combined with the virtual impedance parameters determined in the previous steps to construct a time-varying impedance. Here, the time-varying impedance is not a static resistance value, but a physical constraint mechanism that changes dynamically in the time domain. In order to accurately implement this mechanism at the physical level, the present invention converts the time-varying impedance function into the voltage drop it generates in the circuit for calculation. Specifically, the voltage drop consists of two parts, the first part being the virtual resistance. and virtual inductance For instantaneous current The fundamental impedance voltage drop generated by the response is This term reflects the suppression characteristic of virtual inductance against current surges. The second part is the time-varying modulation term introduced by the binary pseudo-random M-sequence. Combining the two, we obtain the total virtual voltage drop generated by the time-varying impedance at the current moment; Subtracting the total virtual voltage drop from the original driving voltage of the target node yields the actual driving voltage containing the physical watermark. This calculation method effectively avoids the mathematical definition ambiguity that may result from direct nonlinear transformation between the Laplace domain (s-domain) and the time domain (t-domain), ensuring the executability of time-varying impedance in the physical controller. The binary pseudo-random M-sequence is used to construct a physical feature with unique identifiability, serving as a digital key for actively detecting watermarks and marking physical signals in the future. The virtual impedance transfer function calculated in the previous steps is used as the carrier reference, and the binary pseudo-random M-sequence is used as the modulation signal to construct a time-varying dynamic impedance function. Through the time-varying impedance function, instead of outputting a constant impedance value, a dynamic impedance that rapidly jumps between the reference value and the sum of the reference value and the perturbation value with the binary pseudo-random M-sequence. This jump couples the characteristics of the binary pseudo-random M-sequence in the digital domain to the impedance parameters in the physical domain. The calculated virtual voltage drop is subtracted from the drive voltage value in the original control command sent by the target node to obtain the final drive voltage value of the actual drive command. The duration correction is determined by the state deviation. If the state deviation is close to 0, the correction degree is close to 0. If the state deviation is large, the duration will be rapidly compressed by the denominator, thereby achieving rapid blocking at the logical level.

[0032] This invention constructs a time-varying impedance with active defense capabilities by modulating a binary pseudo-random M-sequence into a virtual impedance. By embedding a tiny and uniquely identifiable physical ripple (physical watermark) into the actual driving voltage, the ordinary power drive signal is transformed into a carrier signal with detection capabilities. The duration of the command is dynamically compressed based on the state deviation, further limiting the execution window of suspicious commands. At the physical level, this provides an immutable signal source for subsequent authenticity identification without affecting normal lighting functions.

[0033] Step S3, the process of updating the correlation coefficient based on the target node's response to the physical watermark, specifically includes: The working voltage instantaneous value sequence of the target node within a preset time window is extracted and constructed into a working voltage instantaneous value vector. The binary pseudo-random M sequence and the working voltage instantaneous value vector are subjected to binarization polarity processing to obtain the local polarity reference sequence and the target polarity measurement sequence. The polar cross-correlation function between the local polarity reference sequence and the target polarity measurement sequence is calculated using logical operations. The mathematical expression for the polarity cross-correlation function is as follows: ; in, The value of the polar cross-correlation function. This represents the total number of sampling points within the time window. For sliding time delay, Let k be the value of the binary pseudo-random M-sequence at time k. Let be the voltage at the target node at time k+τ. This is a symbolic function (used to convert data into binary data).

[0034] In a specific embodiment of the present invention, the execution node uses the actual drive command generated in the preceding steps to control its power drive voltage. Since the actual drive voltage contains tiny fluctuations modulated by a binary pseudo-random M sequence, according to Ohm's law and the impedance characteristics of the power supply line, the change in load current will generate voltage physical ripple of the same frequency in the local power supply network. The frequency characteristics of the voltage physical ripple are consistent with the symbol rate of the binary pseudo-random M-sequence, and its amplitude is controlled by the modulation depth coefficient, which is sufficient to be sensed by sensors in the same network segment without affecting the normal luminous stability of the lighting equipment. The binary pseudo-random M-sequence and the instantaneous value vector of the working voltage are converted into a local polarity reference sequence and a target polarity measurement sequence, respectively, using a sign function; The originally complex analog waveform was transformed into two bit streams containing only 0s and 1s, removing amplitude interference and retaining only the phase and frequency characteristics of the fluctuation; By using logical XNOR operations instead of traditional multiplication, the similarity between the local polarity reference sequence and the target polarity measurement sequence is calculated, along with the polarity cross-correlation function. The mathematical expression is: ; in, The value of the polarity cross-correlation function represents the time delay between the locally transmitted signal and the target received signal. The waveform similarity at time, whose value range is normalized to be The closer the value is to 1, the more synchronized the two waveforms are; the closer the value is to 0.5, the less related the two are.

[0035] The total number of sampling points (50-200 in this embodiment) determines the length of the verification time window. Sliding correlation calculations are used to prevent array out-of-bounds operations. The actual number of sampling points must be greater than the effective length of the polarity correlation operation, i.e., the length of the binary pseudo-random M-sequence. This is a sliding variable used to compensate for communication transmission delay and sensor sampling clock deviation, within a preset range (in this embodiment, it is...). Slide within (sampling points) Find the best matching point; Preset sliding search range The actual length of the target node voltage sequence is at least 10 points longer than the length of the binary pseudo-random M-sequence, ensuring that real data is always available during the sliding process. Linear feedback shift registers in The original binary value (0 or 1) generated at any given time. (Symmetric quantization operator) is a mapping function used to map continuous analog quantities to discrete binary quantities; Feedback from the target node at a specific time point The collected raw environmental feature data (i.e., instantaneous feature vectors); Traversing different Value ([-5,+5]) calculation And extract its maximum value as the final physical coupling degree. This directly reflects whether the target node is physically connected in the power supply circuit of the execution node, thus providing quantitative physical evidence for the determination of authenticity in subsequent steps.

[0036] This invention employs binary polarity processing and XOR logic operations to calculate the polarity cross-correlation function. By focusing on the polarity changes of the signal, it can robustly extract weak M-sequence features from noisy power line environments. This calculation method greatly reduces computational complexity, enabling lighting nodes to complete physical watermark detection in real time even with limited computing power, thus ensuring the feasibility of the verification mechanism on low-cost IoT devices.

[0037] In step S3, the process of updating the correlation coefficient based on the target node's response to the physical watermark further includes: If the peak value of the polar cross-correlation function is greater than the preset second correlation threshold, the target node is determined to be a physical connection node; If the peak value of the polar cross-correlation function is less than or equal to the second correlation threshold, the target node is determined to be a virtual deception node.

[0038] In a specific embodiment of the present invention, the preset correlation threshold is 0.8. If the peak value of the polar cross-correlation function is detected to be higher than the preset first correlation threshold, it is determined that the multi-dimensional state data fed back by the target node contains significant and completely synchronized binary pseudo-random M-sequence physical ripple, indicating that the target node is in the same power supply circuit as the execution node in electrical and physical terms and has real perception capability. Therefore, its identity is confirmed as a physical connection node (legitimate node). If the peak value of the polar cross-correlation function is less than or equal to the preset second correlation threshold, it is determined that the data fed back by the target node lacks physical ripple characteristics. This indicates that the target node may be an off-line simulator, a replay attacker, or a faulty device. Therefore, its identity is confirmed as a virtual deception node (illegal node).

[0039] This invention, by setting a correlation threshold, can clearly distinguish whether a target node is a real physical connection node or a virtual deception node that only exists at the network layer. As long as the target node is not in the same power supply circuit, or is an attack node simulated by software, it cannot feed back a voltage signal containing the correct physical ripple, thus being accurately identified and isolated. This fills the gap in traditional network security, which relies solely on cryptographic verification and ignores the authenticity of the physical connection.

[0040] In step S3, the process of updating the correlation coefficient based on the target node's response to the physical watermark further includes: Different preset trust penalty coefficients are assigned to physical connection nodes and virtual deception nodes respectively, and the relevant coefficients are updated in the next round of collaborative control cycle; The mathematical expression for updating the correlation coefficient is: ; in, The original Pearson coefficients are obtained from the multi-dimensional state data collected during the next round of coordinated control. The trust penalty coefficient (in this embodiment, the value is 0 when it is determined to be a virtual node and 1 when it is determined to be a physical node). This is the updated correlation coefficient.

[0041] In one specific embodiment of the present invention, in order to transform the above-mentioned qualitative conclusions about physical connection nodes and virtual spoofing nodes into quantitative mathematical parameters, a trust penalty coefficient is defined. , When the target node is determined to be a physically connected node, let When the target node is determined to be a virtual deception node, then... (Complete blocking) This non-linear penalty mechanism ensures that the trust level of the target node will drop once physical verification fails. The generated trust penalty coefficient is used as prior knowledge and fed back into the next control cycle to update the real-time relevance index (i.e., the original Pearson coefficient). Let the next moment be... The original Pearson coefficient, calculated based on multi-dimensional state data, is: Perform update operation and obtain As The correlation coefficient at time t is used to calculate the subsequent weighted connection weights.

[0042] This invention introduces a dynamic feedback mechanism based on a trust penalty coefficient. Once a target node is detected as a virtual deception node, its weight in the next round of collaborative control is directly reduced by the trust penalty coefficient (e.g., set to zero). This cross-cycle feedback update mechanism can automatically reduce the influence of untrusted nodes on the network over time, gradually forming a safer and cleaner collaborative control environment, effectively preventing repeated malicious attacks.

[0043] A multi-node lighting equipment collaborative control system includes a data acquisition module, a status characteristic module, a virtual impedance module, and a control module; The acquisition module is used to acquire control commands from the target node and collect multi-dimensional status data from the target node, the execution node, and adjacent nodes. The state feature module calculates the consistency of multi-dimensional state data of the target node, execution node, and adjacent nodes as the correlation coefficient, and calculates the state deviation. The virtual impedance module calculates virtual impedance based on state deviation, constructs a pseudo-random sequence, and couples the pseudo-random sequence with the virtual impedance to construct a time-varying impedance. The control module uses time-varying impedance to adjust the control commands of the target node, generates actual drive commands with physical watermarks to drive the execution node, and updates the correlation coefficients based on the target node's response to the physical watermark.

[0044] This invention utilizes a consensus differential equation to calculate the multi-dimensional state deviation between the target node, the execution node, and adjacent nodes, and dynamically maps this deviation to a second-order virtual impedance parameter. The exponentially increasing virtual resistance can effectively attenuate high-amplitude abnormal commands caused by sensor distortion or falsified data, while the virtual inductance, which responds to the rate of change of deviation, utilizes the characteristic of inductance to pass DC and block AC to smooth and suppress pulse-like sudden attack commands. This mechanism of converting data deviation at the information layer into impedance constraints at the physical layer effectively prevents equipment overload or energy waste caused by erroneous strategies. This invention couples a binary pseudo-random M-sequence with a time-varying impedance function to modulate a uniquely identifiable physical watermark (voltage ripple) in the actual driving voltage. It then uses a polarity cross-correlation function to calculate the waveform similarity between the local reference sequence and the target node feedback voltage sequence. This allows for a strict distinction between real physical connection nodes and virtual deception nodes that only exist at the logic layer, going beyond the data appearance. This fills the gap in traditional collaborative control, which relies solely on logic data verification, and ensures the authenticity of the control command execution environment. Based on the response of physical watermarks, this invention uses a trust penalty coefficient to apply a nonlinear penalty to the association weight of deceitful nodes, and updates the correlation coefficient in the next round of collaborative control. This closed-loop feedback mechanism can automatically isolate untrusted nodes as the running time progresses, thereby ensuring the high reliability and anti-interference capability of the multi-node lighting equipment collaborative control network in long-term operation.

[0045] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0046] It should be noted that the above 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 preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the protection scope of the present invention.

Claims

1. A method for coordinated control of multi-node lighting devices, characterized in that, Includes the following steps: Step S1: Collect control commands from the target node and collect multi-dimensional status data of the target node, execution node, and adjacent nodes; Step S2: Calculate the correlation coefficient between the multi-dimensional state data of the target node, the execution node and the neighboring nodes. Based on the multi-dimensional state data of the neighboring nodes, filter the effective dimensions in the multi-dimensional state data. Substitute the effective dimension state data of the target node into the consistency differential equation weighted by the correlation coefficient and iterate until convergence. Calculate the state deviation. Step S3: Construct virtual resistance parameters based on the amplitude of the state deviation, construct virtual inductance parameters based on the rate of change of the state deviation over time, the virtual resistance parameters and virtual inductance parameters constitute virtual impedance, construct a pseudo-random sequence, and couple the pseudo-random sequence with the virtual impedance to construct a time-varying impedance; Step S4: Adjust the control command of the target node using the time-varying impedance, generate an actual drive command with a physical watermark to drive the execution node, and update the correlation coefficient according to the target node's response to the physical watermark.

2. The multi-node lighting equipment collaborative control method as described in claim 1, characterized in that, In step S1, the target node is the source lighting device node that initiates the control command; The execution node is a lighting device node that receives the control command and is responsible for performing physical lighting actions; The adjacent nodes are other lighting device nodes that have a direct physical communication connection with the execution node; The control command includes the drive voltage and duration; The multi-dimensional state data includes voltage-dimensional state data, current-dimensional state data, ambient light-dimensional state data, moving object detection-dimensional state data, and moving object speed-dimensional state data. The voltage dimension state data is a sequence of instantaneous working voltage values ​​obtained by sorting the instantaneous working voltage values ​​by time within a time window; The current dimension state data is a sequence of instantaneous operating current values ​​obtained by sorting the operating current values ​​within a time window by time. The ambient light intensity dimension status data is an ambient light intensity sequence obtained by sorting the ambient light intensity by time within a time window; The moving object detection dimension state data is a sequence of moving object detection states obtained by sorting the moving object detection states by time within a time window. The state data for the velocity dimension of the moving object is a velocity sequence obtained by sorting the detected velocities of the moving object within a time window by time.

3. The multi-node lighting equipment collaborative control method as described in claim 2, characterized in that, In step S2, the process of calculating the consistency of the multi-dimensional state data of the target node, the execution node, and the adjacent nodes specifically includes: Extract multi-dimensional state data from the target node, execution node, and adjacent nodes respectively; The state data of the target node in each dimension are the state data of each dimension within a preset time window before the control command is generated; The status data of each dimension of the execution node are the status data of each dimension within a preset time window before receiving the control command from the target node; The state data of each dimension of the adjacent node is the state data of each dimension within a preset time window before the executing node receives the control command from the target node. Calculate the correlation coefficient between the state data of adjacent nodes and the execution node in each dimension; If the correlation coefficient between the state data of adjacent nodes and execution nodes is greater than the preset first correlation threshold, the corresponding dimension is determined to be a valid dimension, and the state data of the valid dimensions of the target node and execution node are constructed into a valid state set.

4. The multi-node lighting equipment collaborative control method as described in claim 3, characterized in that, In step S2, the process of calculating the consistency of the state data of the target node, the execution node, and the adjacent nodes further includes: Within the set of valid states, the correlation coefficients of the multi-dimensional state data of the target node and the multi-dimensional state data of the execution node are calculated respectively. The target node and its adjacent nodes are used as connection nodes for the execution node. The connection weights between the execution node and the connection nodes are constructed, and the initial value of each connection weight is 1. Using the correlation coefficients of execution nodes and connected nodes under the effective dimension, the connection weights are weighted and corrected to obtain the weighted connection weights. The mathematical expression for the weighted connection weights is as follows: ； in, For the corrected connection weights, The initial values ​​for the connection weights are... For execution node With the target node and neighboring nodes The correlation coefficient; The multi-dimensional state data of the target node, execution node, and adjacent nodes at a certain moment are vertically concatenated into a column vector in the effective dimension to obtain the instantaneous state vector at that moment. The instantaneous state vector is then substituted into a first-order consistency differential equation for consistency verification. The mathematical expression of the consistency differential equation is as follows: ； in, The change in the state data of the execution node. It is the set of neighboring nodes. Let be the column vector of the neighboring nodes at time t. This is the column vector of the execution node at time t; Iterative calculations are performed until convergence, yielding consistent convergence state data. The state deviation is obtained by calculating the Euclidean distance between the state data of the target node and the consistent converged state data.

5. The multi-node lighting equipment collaborative control method as described in claim 4, characterized in that, In step S2, the process of constructing the virtual impedance specifically includes: A virtual impedance is constructed based on the state deviation, and the virtual impedance includes virtual resistance parameters and virtual inductance parameters. The mathematical expression for the virtual resistance parameter is: ； in, For virtual resistance, Based on the fundamental resistivity, This is the resistance gain factor. This refers to the degree of state deviation. The virtual inductance parameters : ； in, Based on the fundamental inductance coefficient, The rate of change of the deviation from the state over time. This represents the degree of state deviation.

6. The multi-node lighting equipment collaborative control method as described in claim 5, characterized in that, In step S3, the process of adjusting the control command of the target node specifically includes adjusting the driving voltage and duration, and obtaining the actual driving voltage and actual duration. A binary pseudo-random M-sequence within a preset time window is generated using a linear feedback shift register, and the binary pseudo-random M-sequence is used as a modulation factor and coupled with a virtual impedance to form a time-varying impedance. The actual driving voltage is obtained by calculating the virtual voltage drop generated by the time-varying impedance in the execution node circuit. The mathematical expression for the actual driving voltage is as follows: ； in, This is the actual driving voltage. The drive voltage value in the control command sent to the target node. For virtual resistance parameters, For virtual inductance parameters, For the execution node at time Instantaneous operating current, This is the modulation depth coefficient (in this embodiment, the value range is 0.05~0.1). The mathematical expression for the actual duration is: ； in, This refers to the actual execution window duration. The time sensitivity coefficient (in this embodiment, the value is 10). To control the duration in the command, This represents the degree of state deviation.

7. The multi-node lighting equipment collaborative control method as described in claim 6, characterized in that, In step S3, the process of updating the correlation coefficient based on the target node's response to the physical watermark specifically includes: The working voltage instantaneous value sequence of the target node within a preset time window is extracted and constructed into a working voltage instantaneous value vector. The binary pseudo-random M sequence and the working voltage instantaneous value vector are subjected to binarization polarity processing to obtain the local polarity reference sequence and the target polarity measurement sequence. The polarity cross-correlation function between the local polarity reference sequence and the target polarity measurement sequence is calculated using logical operations. The mathematical expression of the polarity cross-correlation function is as follows: ； in, The value of the polar cross-correlation function. This represents the total number of sampling points within the time window. For sliding time delay, Let k be the value of the binary pseudo-random M-sequence at time k. Let be the voltage of the target node at time k+τ.

8. The multi-node lighting equipment collaborative control method as described in claim 7, characterized in that, In step S3, the process of updating the correlation coefficient based on the target node's response to the physical watermark further includes: If the peak value of the polar cross-correlation function is greater than the preset second correlation threshold, the target node is determined to be a physical connection node; If the peak value of the polar cross-correlation function is less than or equal to the second correlation threshold, the target node is determined to be a virtual deception node.

9. The multi-node lighting equipment collaborative control method as described in claim 8, characterized in that, In step S3, the process of updating the correlation coefficient based on the target node's response to the physical watermark further includes: Different preset trust penalty coefficients are assigned to physical connection nodes and virtual deception nodes respectively, and the relevant coefficients are updated in the next round of collaborative control cycle; The updated mathematical expression for the correlation coefficient is as follows: ； in, The original Pearson coefficients are obtained from the multi-dimensional state data collected during the next round of coordinated control. The trust penalty coefficient, This is the updated correlation coefficient.

10. A multi-node lighting equipment collaborative control system, applied in the multi-node lighting equipment collaborative control method as described in any one of claims 1-9, characterized in that, It includes an acquisition module, a status characteristic module, a virtual impedance module, and a control module; The acquisition module is used to acquire control commands from the target node and multi-dimensional status data of the target node, the execution node, and adjacent nodes. The state feature module calculates the consistency of multi-dimensional state data of the target node, execution node, and adjacent nodes as a correlation coefficient, and calculates the state deviation. The virtual impedance module calculates virtual impedance based on the state deviation, constructs a pseudo-random sequence, and couples the pseudo-random sequence with the virtual impedance to construct a time-varying impedance. The control module uses the time-varying impedance to adjust the control commands of the target node, generates actual drive commands with physical watermarks to drive the execution node, and updates the correlation coefficients according to the target node's response to the physical watermarks.

Images