High geological disaster chain monitoring node deployment method based on wireless ad hoc network

Abstract

The invention discloses a high-level geological disaster chain monitoring node deployment method based on a wireless ad hoc network, and relates to the technical field of geological disaster monitoring and early warning, and the method comprises the following steps: S1, obtaining interference characteristics through a time-frequency observation layer triggered by an earthquake, and building a time baseline; s2, identifying a cross-channel coupling relationship according to the energy spectrum and the drift baseline, and determining sensitive nodes; s3, reconstructing a convergent path by replacing a distortion channel with a shadow link; s4, constructing a time stack and suppressing routing oscillation by a hysteresis gate; s5, generating a time-varying virtual impedance route according to the time stack to realize peak shifting stable forwarding; and S6, completing broken-loop control of residual crosstalk through inverse phase traction and the metasurface antenna, and finally enabling the network to maintain long-term stable operation. According to the method, stable convergence and closed-loop regulation and control of the ad hoc network are realized through interference feature extraction, path reconstruction, time stack hysteresis vibration suppression, virtual impedance peak shifting forwarding and inverse phase broken-loop interference control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geological disaster monitoring and early warning, and particularly relates to a high-position geological disaster chain monitoring node deployment method based on a wireless ad hoc network. BACKGROUND

[0002] The "high-position geological disaster chain monitoring node deployment based on a wireless ad hoc network" refers to the following: in a high-position geological disaster chain area (such as a mountain landslide, collapse, debris flow, etc. possible occurrence zone), a plurality of dispersedly arranged monitoring nodes are organically connected according to the spatial distribution of the disaster chain, an edge gateway is taken as a core, and signal interconnection and intercommunication between nodes and between the nodes and the gateway are realized through a relay mode. In this deployment mode, the monitoring nodes can not only actively trigger and upload on-site data under abnormal conditions, but also passively feedback according to the instructions of the edge gateway, so as to realize two-way data interaction and flexible control. Meanwhile, through the design of a wireless ad hoc network protocol with a priority scheduling mechanism, the monitoring information can be transmitted in time periods and in order of urgency according to the different positions and risk levels of the disaster chain, so as to ensure that important data is preferentially arrived and to improve the real-time performance, stability and reliability of the overall monitoring. This deployment method not only solves the problems of wiring difficulty and signal shielding under high-position complex topography, but also can construct a self-adaptive and expandable disaster chain monitoring system to provide efficient data support for disaster early warning.

[0003] The prior art has the following disadvantages: In the prior art, when the monitoring nodes are deployed in the high-position geological disaster chain area and are affected by aftershocks, the radio frequency circuit of the nodes is extremely easy to produce electromagnetic coupling crosstalk due to strong vibration and environmental disturbance, so that interference occurs between originally independent signal channels, and then abnormal distortion of the communication signal is caused. In this case, the ad hoc network protocol automatically triggers topology reconstruction to maintain link connectivity, but due to the cumulative effect of the interference, a routing loop is easily formed in the reconstruction process, data is repeatedly forwarded in the loop and cannot reach the edge gateway, and finally the loop lock phenomenon is caused. Once this phenomenon occurs, the entire ad hoc network will be in a non-convergent state for a long time, the monitoring data cannot be stably transmitted, and the real-time monitoring and early warning of the disaster chain are seriously hindered.

[0004] The above information disclosed in the background section is only used to enhance the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0005] The purpose of the present application is to provide a high-position geological disaster chain monitoring node deployment method based on a wireless ad hoc network to solve the problems in the background.

[0006] In order to achieve the above object, the present application provides the following technical scheme: A high geological disaster chain monitoring node deployment method based on wireless ad hoc network, comprising the following steps: S1, constructing a time-frequency observation layer triggered by an earthquake in a high geological disaster chain area, laying a double-path coupling probe and introducing a phase calibration beacon, collecting dynamic electromagnetic interference field information, extracting crosstalk energy spectrum and time drift baseline as a reference basis for subsequent processing; S2, based on the crosstalk energy spectrum and the time drift baseline, performing causal coherence decomposition, identifying cross-channel coupling kernels and analyzing energy propagation direction, and calibrating a sensitive node set causing routing anomalies; S3, constructing an anti-factual playback chain based on the sensitive node set, replacing the signal distortion segment with a shadow link, reconstructing a convergent routing track, and restoring the time sequence consistency between nodes; S4, deploying a double-mirror time tag around the routing track, constructing a unified time stack, and injecting a hysteresis gate into the time stack to suppress oscillation caused by loop backflow and topology reconstruction; S5, generating a time-varying virtual impedance route based on the time stack, constructing a topology potential energy landscape, outputting a node forwarding sequence with peak-shifting characteristics, and guiding the network into a stable convergence state; S6, after the network enters a stable convergence state, performing time reversal phase traction, driving a programmable metasurface antenna array by injecting inverse phase micro-pulses, and combining a shadow energy storage structure to absorb residual crosstalk energy, completing the loop breaking control of coherent topology, and realizing closed-loop dynamic regulation.

[0007] Preferably, step S1 comprises: Selecting a typical section with landslide, collapse or debris flow induction risk in a high geological disaster chain area, determining the layout boundary in combination with geological exploration results and terrain modeling data; Laying double-path coupling probes at equal intervals in the layout area, each coupling probe collecting disturbance electric signals in the ground and on the ground, and connecting to a high-sensitivity sampling terminal through independent wires; Installing a group of phase calibration beacons around each probe, the beacons actively emitting reference pulses with constant amplitude and frequency and consistent phase after an earthquake, and synchronously injecting adjacent probes to form a phase reference; After an earthquake disturbance occurs, collecting original disturbance waveforms of each channel, extracting amplitude variation, phase shift and main peak energy density, and generating a crosstalk energy spectrum and a time drift baseline.

[0008] Preferably, step S2 comprises: According to the crosstalk energy spectrum and the time drift baseline, constructing an interference time chain and generating a multi-node influence path map; On the basis of path atlas, the propagation level is divided, the frequency domain features are extracted, and the node pairs with stable phase delay and high coherent amplitude are identified to construct a cross-channel coupling kernel database; Based on the coupling kernel database, the propagation path network is constructed and the energy propagation atlas is generated, and the energy vector field is formed combined with the three-dimensional geographic coordinates; In the energy vector field, the path aggregation point and the disturbance mutation point are identified, the path shock response of the node is verified by the disturbance simulation playback test method, and finally the sensitive node set is determined.

[0009] Preferably, step S3 comprises: According to the sensitive node set and the distortion channel segment list, the transmission abnormal path is identified, and the corresponding shadow link is constructed to replace the original path segment; The shadow link path is determined by hop number control, spatial boundary restriction, signal quality evaluation and backhaul integrity rate screening; With the time drift baseline as the reference, the node timing reconstruction is performed, and the time alignment and error calibration of all nodes in the shadow link are performed; Through the path integration operation, the shadow link is connected to the main path architecture, and the signal forwarding suppression mechanism and routing table redirection control are applied to realize path convergence and stable operation.

[0010] Preferably, the selection of the shadow link path needs to meet the requirements that the hop number is not more than one and a half times of the original path hop number, the path spatial range is limited within fifty meters of the original path, and the signal-to-noise ratio of all relay nodes is higher than twenty-five decibels, and the signal backhaul integrity rate is not less than ninety-five percent.

[0011] Preferably, step S4 comprises: Double-mirror time marker devices are deployed at both ends of the reconstructed routing trajectory starting node and receiving node to complete time synchronization and direction verification between path hop points; Time marker data of all path nodes are collected to construct a time stack structure and calibrate the propagation delay and interference sensitivity coefficient to generate a routing time sequence matrix; The path segment with nonlinear growth characteristics in the time stack is identified, and a hysteresis gate with adjustable response delay is injected to form a forwarding buffer and path freezing mechanism; Multiple rounds of path transmission tests are performed to verify the time marker synchronization, gate delay control, and path topology oscillation suppression effect to ensure the controllability and stability of the time stack.

[0012] Preferably, the injection position of the hysteresis gate is limited to the position where the propagation delay increase is more than twice the previous level and is accompanied by a sudden drop in signal energy and a phase disturbance peak, so as to ensure that the path oscillation source is accurately constrained.

[0013] Preferably, step S5 comprises: Extract the average propagation delay of each hop path node in the time stack, generate a node time impedance table, and build a multi-cycle time-varying virtual impedance trajectory; Superimpose each node impedance trajectory on the historical energy distribution to form a topological barrier map with spatial continuity and time fluctuation perception, and label the barrier nodes and low impedance regions; According to the impedance difference, divide the path section, make a segmented peak-shifting forwarding scheduling plan, trigger node forwarding in sequence according to impedance priority, and generate a path-level forwarding sequence plan; Set a sub-section convergence verification window, evaluate whether the data transmission is completed according to the scheduling, adjust the forwarding frequency and timestamp according to the feedback result, and finally realize the coordinated convergence of the whole network.

[0014] Preferably, in the segmented peak-shifting forwarding scheduling plan, the interval of node forwarding trigger time of each sub-section is set to be a preset multiple of the maximum propagation delay in the section, so as to avoid overlapping of node forwarding, and when any sub-section does not meet the convergence verification condition for two consecutive rounds, the path is marked as a jitter path section, and a backup path replacement mechanism is started to maintain the sustained and stable convergence of the whole network.

[0015] Preferably, step S6 comprises: Identify nodes in the communication path that have asymmetric phase inversion and delay shrinkage, extract the reverse propagation time period and establish a reverse intervention target list; Inject an inverse phase micro-pulse signal into the intervention target node, propagate in reverse order at the tail of the path according to the time stack order, and detect whether there is a persistent phase reconstruction response in the path; Deploy a programmable metasurface antenna array around the detection node, and through the synergistic action of the phase-controlled reflection unit and the shadow energy storage structure, block the loop path in the physical space; According to the traction result, adjust the forwarding priority and shield the risk direction, ensure that the whole network topology has only one path for forward propagation, and form a dynamic regulation and control structure of closed loop broken loop.

[0016] In the above technical solutions, the technical effects and advantages provided by the present application are: The present application realizes accurate modeling of physical interference sources by constructing a time-frequency observation layer triggered by seismic events, collecting electromagnetic interference features in real time and extracting crosstalk energy spectrum and time drift baseline; subsequently, using causal coherence decomposition and counterfactual playback chain technology, actively identifying and repairing distorted node links in the communication path, ensuring that the routing trajectory has stable convergence capability; further, by deploying double-mirror time scales and hysteresis gates to construct a unified time stack structure, suppressing loop backflow and signal oscillation in topology reconstruction; on this basis, generating time-varying virtual impedance routing with segmented staggered peak characteristics to guide the network to operate stably segment by segment; finally, with the programmable metasurface antenna array driven by inverse phase micro-pulse and shadow energy storage structure, active identification and physical ring breaking control of residual crosstalk paths are realized, and a coherent topology structure with long-term evolutionary adaptability and closed-loop control capability is constructed. This method improves the anti-interference, adaptability and topology stability of wireless ad hoc networks in high geological disaster chain areas as a whole, and provides a solid technical support for high-frequency and high-precision disaster monitoring and early warning. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described below are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0018] Figure 1 The method flow chart of the present application is a method for deploying a high geological disaster chain monitoring node based on a wireless ad hoc network. DETAILED DESCRIPTION

[0019] Example implementations will now be described more fully with reference to the accompanying drawings. Example implementations may, however, be implemented in many different forms and should not be construed as limited to the examples set forth herein; rather, these example implementations are provided so that this disclosure will be thorough and complete, and will fully convey the inventive concept to those skilled in the art.

[0020] The present application provides a method for deploying a high geological disaster chain monitoring node based on a wireless ad hoc network as shown in Figure 1 The method flow chart of the present application is a method for deploying a high geological disaster chain monitoring node based on a wireless ad hoc network. S1, constructing a time-frequency observation layer triggered by seismic events in a high geological disaster chain area, laying double-coupling probes and introducing phase calibration beacons, which are used to collect dynamic electromagnetic interference field information when a seismic disturbance occurs, and extract crosstalk energy spectrum and time drift baseline, forming a reference for subsequent signal processing; In the high geological disaster chain area where earthquakes occur frequently, the method of constructing time-frequency interference characteristics is used for monitoring node deployment, and the specific process is as follows: In the high geological disaster chain area, typical sections with landslide, collapse or debris flow induced risk are selected, and the layout boundary is determined according to the geological exploration report and the three-dimensional terrain modeling results. The layout area should cover the fault slip area, the slope top crack zone, the valley confluence and the position of rock and soil exposed by wind and erosion, and combined with historical earthquake damage data, the fissure enrichment zone with high seismic sensitivity is preferentially selected. Along these key parts, double-path coupled probes are arranged at equal intervals, the probes adopt an insulating ceramic shell and a platinum electrode contact end, one end is embedded in the rock body for 10 to 15 cm, and the other end is exposed within 10 cm of the surface to directly contact the air, and the disturbed electric signals in the ground and on the ground are collected respectively. The tail of the probe is provided with a shielding layer connection line, which is led into an independent high-sensitivity sampling terminal. In order to build a stable phase reference, a set of phase calibration beacons is installed within 3 meters around each probe, which includes a high-frequency resonant cavity, a piezoelectric excitation source and a vibration initiation pulse generator. After receiving the foreshock signal, it automatically activates, emits a reference pulse wave train with constant amplitude and frequency, consistent phase and unique direction, and synchronously injects it into the adjacent probe through the coupling capacitor path, to realize instantaneous waveform alignment and phase reference establishment. The starting threshold of the beacon is set to be greater than 0.78 m / s² (equivalent to about 0.08 g) of vertical acceleration, to ensure that it responds only in vibration events with engineering disturbance significance.

[0021] After the layout of the coupled probe and the phase calibration beacon is completed, the seismic disturbance data collection phase is entered. Each probe signal is connected to an analog-to-digital conversion acquisition unit with 16-bit resolution in an independent channel, and the sampling rate is set to 5000 times per second. After the vibration event is triggered, all channels enter the parallel capture state, and take the reference pulse first emitted by the beacon as the unified starting point, and continuously record the original disturbance waveform for not less than 10 seconds before and after. Based on the difference track of the two channel signals after preliminary processing, the amplitude variation, phase shift and corresponding frequency band main peak energy density per unit time are extracted. By aligning and analyzing the phase change curves of different probes at the same time, the interference coupling relationship matrix between them is established, and a preliminary crosstalk energy spectrum diagram is generated according to the coupling strength. In order to accurately calibrate the response delay of different nodes, the time axis of all signals is further adjusted to the zero point corresponding to the beacon pulse, and through trend fitting of the small drift in different time periods, a four-dimensional time drift baseline table containing probe number, response delay, phase shift value and signal similarity is generated, which is used to identify the dynamic changes of the disturbance propagation link.

[0022] Based on the generated crosstalk energy spectrum and time drift baseline table, the interference signal is cleaned, classified and evolution path is reconstructed. First, the high frequency signal within 3 seconds before and after the earthquake triggering time is compared with the background noise band point by point, and the non-seismic disturbance signal is removed. Then, according to the concentration degree of energy distribution in the energy spectrum, the spectrum of each probe is divided into multiple fixed frequency bands, and the energy change rate and phase disturbance of each frequency band before, during and after the vibration are extracted. These disturbance characteristics are mapped into frequency domain transition track, and combined with spatial geographic coordinates, a complete three-dimensional interference evolution model is formed. By marking the phase inversion node, coupling strength peak path and its propagation direction in the model, the transfer process of disturbance energy between probes is gradually revealed, and finally an interference propagation trend table is generated, which is composed of probe number, starting disturbance time, phase inversion time point, maximum coupling direction and corresponding energy value. The trend table clearly points out the node position where the signal distortion may start and its outward radiation path, providing an accurate basis for subsequent fault routing identification.

[0023] Based on the energy aggregation area reflected by the interference propagation trend table, combined with the time drift baseline table established before, the response synchronization analysis of all disturbed nodes is carried out. Taking the first disturbance time of each node as the reference point, the average response delay of each node and the remaining probes is calculated, and if the delay exceeds the set threshold of 0.5 milliseconds, it is determined as a response inconsistent node. For the probes with consistent response but sudden change in phase change waveform, they are further judged as high-risk disturbance sources. The above identified nodes are combined to form a high sensitivity disturbance point set, and an interference influence weight matrix is constructed according to the spatial distance and time response difference between them. Then, two-dimensional projection is performed on the matrix to form a standard interference mapping diagram containing three dimensions of time, space and energy. Finally, taking the interference mapping diagram as input, a "time-frequency disturbance response reference layer" with spatial continuity, time progression and energy visibility is derived, which not only has the characteristics of actual seismic disturbance field, but also has a quantitative reference for coupling identification. It is an important input data source for subsequent causal coherence decomposition and sensitive node calibration operations.

[0024] S2, based on the extracted crosstalk energy spectrum and time drift baseline, performing causal coherence decomposition, identifying cross-channel coupling cores between monitoring nodes layer by layer, analyzing energy propagation direction, and marking sensitive node set prone to routing abnormality; After obtaining the crosstalk energy spectrum and time drift baseline table with time alignment accuracy and frequency spectrum stability, in order to identify the cross-channel coupling relationship between disaster chain monitoring nodes and the sensitive node position, the following operations are implemented: According to the constructed interference propagation trend table and the time drift baseline table, the interference starting nodes with concentrated response time and significant spectrum mutation are screened out, and are taken as the starting reference of the causal path. Taking these starting nodes as the center, interference time chains are constructed to all surrounding nodes, and the time chain is composed of specific indicators such as signal starting response time, phase rise slope, spectrum jump position and energy concentration section. Each chain represents a possible interference transmission relationship between node pairs, and the construction of the time chain is based on the sliding window sequence point-by-point comparison method, and the window step is set to 0.5 seconds, and the time span is set to 5 seconds before and after the vibration trigger point. The data difference between each group of nodes needs to be standardized by the time drift baseline table to eliminate the offset caused by the difference in local clock accuracy of different nodes. The complete time chain is sorted according to the time sequence from the interference starting point to the response point, and the directionality index is marked, so as to obtain a multi-node influence path map with causal relationship preliminary judgment ability, which provides ordered input for subsequent coupling analysis.

[0025] On the basis of the path map, the interference conduction process is divided into multiple propagation levels according to the time sequence, and each level represents a set of response nodes within a fixed time delay section. The signal amplitude rapid transition point, phase sudden reversal point and spectrum instantaneous collapse point appearing in each propagation layer are combined and paired to construct a set of suspected coupling node pairs. The frequency domain signals of each group of node pairs are decomposed into multiple frequency bands, specifically into 50-150 Hz, 150-300 Hz, 300-600 Hz and 600-1000 Hz four frequency bands, and the cross-correlation amplitude, phase angle change trend and waveform envelope retention of each frequency band are extracted. If there are more than three node pairs with stable phase delay trend and high coherent amplitude (more than 60% of the reference beacon amplitude) in three or more frequency bands, it is determined that there is an actual energy coupling relationship, and it is divided into a cross-channel coupling core. In order to ensure that the identified coupling core has directionality and level awareness, the propagation before and after the level number, signal starting and ending time point, response delay interval and spectrum main peak drift rate corresponding to each group of coupling cores are recorded, and finally constitute a cross-channel coupling core database. The database comprehensively represents the dynamic coupling path distribution characteristics between nodes, and is a basic data source for accurately reflecting the interference propagation law from the signal evolution level.

[0026] After obtaining all the cross-channel coupling kernel data, the coupling kernels are used as nodes to build a propagation path network. The propagation path network is connected by one-way edges, and the edge weight is defined as the phase disturbance amount when a unit of energy flows through it. The direction is from the energy source node to the energy receiving node. Combining the start and end time points of all coupling kernels and the propagation path level number, an energy propagation graph containing a complete direction vector is constructed. The graph is projected into a three-dimensional geographical coordinate model to form a three-dimensional energy migration path graph. Then, based on the propagation path density and node aggregation degree, vector aggregation analysis is performed on the path graph. Specifically, within a 1-meter range around each node, the number of paths pointing to the node is counted. If the number exceeds twice the average value and the path direction concentration is greater than 0.7 (calculated from the direction vector angle), the node is identified as an energy aggregation point. Conversely, if a node has frequent path direction changes, path breaks, and poor energy continuity, it is marked as a disturbance mutation point. The spatial distribution of all aggregation points and mutation points is further mapped into a disturbance energy vector field, from which interference accumulation areas, path turning areas, and coupling kernel dense bands can be identified. These areas often correspond to topological bottlenecks or unstable nodes in network communication.

[0027] By combining the constructed energy vector field, path propagation graph, and coupling kernel database, a set of sensitive nodes with high disturbance feedback potential is identified. To verify the impact of these nodes on the communication topology, a disturbance simulation playback method is used to test the stability of the paths. This test inputs an artificial disturbance signal with an amplitude of 0.98 m / s² (approximately equal to 0.1 g) from the upstream of each candidate sensitive node. The simulation signal is a continuous oscillation waveform lasting 3 seconds, synthesized from the main frequency components of the earthquake waves recorded in the previous real disturbance record. The actual delay, energy attenuation ratio, and whether the path is deformed are recorded as three indicators. The test is repeated three times with independent disturbances. If a node triggers path oscillation, repeated data transmission failure, or overlapping disturbance feedback from adjacent nodes in all three rounds, it is determined to be a structural sensitive node. The final set of sensitive nodes includes all nodes with high coupling strength, high path aggregation degree, and high propagation instability. Their positions and propagation characteristics will directly serve as the basis for the next step of shadow link replacement and route trajectory reconstruction, and will be used to accurately repair the routing loop abnormality problem caused by earthquakes.

[0028] S3, according to the sensitive node set, construct an anti-fact playback chain, use shadow link replacement to identify the signal distortion channel segment, reconstruct the routing trajectory with stable convergence ability, and synchronize the time sequence consistency between the monitoring nodes in the routing trajectory; To solve the node crosstalk and routing loop instability problems caused by earthquakes, after identifying the sensitive node set, it is necessary to reconstruct the routing trajectory with time sequence consistency and stable convergence ability. This is implemented in the following steps: Based on the calibrated sensitive node set and the list of signal distortion channel segments, all distorted path segments are extracted from the original topology, and the communication direction, signal strength variation characteristics and propagation delay anomaly curve of the nodes before and after the path segment are determined. The demarcation of the distorted path segment does not depend on the simple transmission failure judgment, but takes the number of frequency spectrum jumps, the amplitude of phase inversion angle mutation, and the discontinuity of energy distribution as objective evaluation basis during the transmission of the signal on the path, and screens out the channels with continuous transmission anomaly rate higher than 30%, propagation phase jump more than 45 degrees, and path energy decreasing rate more than 70%. For each distorted channel segment, a separate identification is made in the original topology graph to confirm its hop position in the overall network communication path, node distance, terrain shielding information and distance boundary with adjacent nodes, laying a foundation for subsequent construction of shadow link.

[0029] For the path empty segment after removing the distorted segment, a shadow link is designed to achieve equivalent replacement. The construction principle of the shadow link is to find a feasible connection path with the number of hops controlled within 1.5 times of the original path and the geographical path not exceeding the 50-meter range around the distorted path without relying on the distorted node. In order to ensure that the new path has anti-interference in signal quality, the signal attenuation of each hop path is judged comprehensively from the factors such as terrain slope, shielding structure, electromagnetic background noise level. In the specific implementation process, the starting node of the original distorted segment is scanned outward to find a group of candidate relay nodes with delay not exceeding 1 second, average signal-to-noise ratio not less than 25 decibels, and spatial distance not more than 30 meters, and then a group of hop points with stable connection records are selected from the candidate relay nodes. All candidate nodes need to complete not less than 95% signal return integrity in three rounds of test transmission, and then enter the candidate shadow link pool. On this basis, the optimal shadow link path is selected according to the overall time consumption, node transfer frequency and spatial distribution balance of the path, and the propagation parameters, frequency spectrum matching degree and coupling strength index of each hop are recorded. The final determined shadow link is bound with the original path in a one-to-one manner in a hop-by-hop manner, replacing the corresponding communication interval in the original distorted path segment.

[0030] After the deployment of the shadow link physical structure, timing reconstruction needs to be performed for all nodes participating in the communication of the link to restore the time consistency of the entire path. This step refers to the time drift baseline table to calculate the actual transmission delay of each node in the new path under the shadow link, and synchronously converts the timestamps in the original path to the new path to ensure that the starting trigger time and the receiving processing time of all forwarding nodes maintain a linear stable relationship. The local clock of each node is triggered again by the reference beacon signal to calibrate and achieve millisecond-level alignment. Single-hop round-trip transmission verification is performed on all shadow link nodes, and the starting time, arrival time, and confirmation feedback time of the received signal of each node are recorded. The error in the continuous three round-trips should not exceed 0.1 seconds. If there is a deviation fluctuation in a node during the calibration process, the relay priority needs to be lowered or replaced in the structure to prevent it from becoming a potential trigger point for subsequent path divergence. Through multiple rounds of timing regulation, the time sequence consistency between nodes in the new path is finally achieved, ensuring that the shadow link has coherent, continuous, and predictable communication behavior in the logical topology.

[0031] After the structure replacement and timing adjustment are completed, the path integration operation is performed to formally connect the newly built shadow link to the network main path architecture. To prevent the original distorted path from being selected again during the network self-recovery process and causing loop recurrence, a signal forwarding suppression mechanism is applied to all replaced nodes. This mechanism is not implemented by physically disconnecting the link, but by introducing hop count suppression and node state shielding identification during path broadcasting to block its participation in route reconstruction calculation. At the same time, during the route table update process, the shadow link path is set as the priority forwarding path, and all link requests sent to the original path are automatically redirected to the shadow link entry node, ensuring that the new path becomes the preferred transmission channel in actual communication. Three rounds of complete data backhaul testing are performed on the entire path, with an average data packet delay of less than 1 second, a data packet loss rate of less than 5%, and a route reconstruction frequency of zero within the observation period, indicating that the path has stable convergence ability and can be operated for a long time. Finally, the above-mentioned process realizes a closed-loop processing flow from identification, replacement, reconstruction to confirmation, effectively eliminating the transmission link affected by the earthquake disturbance, and constructing a new communication path structure with consistent timing, stable structure, and strong anti-interference ability, providing a solid foundation for subsequent time stack construction and topology regulation.

[0032] S4, deploying double-mirror time markers around the reconstructed route trajectory, mapping the route trajectory to a unified time stack structure, and injecting a hysteresis gate with response delay characteristics into the time stack to suppress loop backflow and limit the oscillation amplitude during topology reconstruction; To further suppress the path loop back and topology reconstruction oscillation problems caused by earthquake disturbance in the network, the following operations need to be performed based on the reconstructed shadow route trajectory: Around the shadow routing track that has been built, a pair of double-mirror time tag devices is deployed at both ends of the starting node and receiving node of each hop path to form a strictly symmetrical time boundary. The time tag device is composed of a high-precision crystal oscillator, a constant-temperature clock driving circuit and a phase calibration charge component, which can maintain millisecond-level frequency stability and time accuracy at different times. In the layout process, the shadow link starting point is taken as the time zero reference point, and the corresponding time tag is configured for all hop points in turn from the starting node of the path in the communication direction. The sending time and receiving time on each hop point are recorded and bidirectionally paired and checked with their mirror nodes, thereby realizing bidirectional correction of time mapping. The time tag starts with a unified pulse emitted by the reference beacon, and after compensation processing of the propagation delay of the path, it continuously aligns all downstream nodes, ensuring that the time transfer process between any two adjacent hop points has continuity, verifiability and directionality, avoiding path reverse misjudgment or link loop error establishment caused by local node clock drift.

[0033] After the double-mirror time tag deployment of all path nodes is completed, the time tag data of each node is collected to build a complete time stack structure. The time stack structure is based on the shadow routing track, and the sending time and receiving time of each node are respectively taken as the in-stack elements of the time stack, which are stacked from the source node to the terminal point in the path order. In the construction process, the time difference between the in-stack position of each node and the out-stack position of the previous node is calculated and set as the propagation delay record value of the stack layer. To improve the response ability of the time stack structure to micro-disturbance, a group of disturbance sensitivity coefficient indicators is added in each stack layer to represent the signal stability, signal-to-noise ratio fluctuation amplitude and coupling strength change trend of the path segment in historical seismic disturbance. Through the stack layer index table, these additional information is bound one by one with the time level to realize the structure stability evaluation function of the time stack. All time data is uploaded to the central control node, which is uniformly summarized to form a complete "routing time sequence matrix". The matrix completely corresponds to the path structure in space and presents linear progressive characteristics in time, with accurate mapping, complete recording and full traceability ability.

[0034] After identifying the sensitive section that may trigger path oscillation in the time stack, a hysteresis gate with a response delay adjustment function is injected into the time stack. The hysteresis gate is a physical timing control device composed of a delay crystal array, a signal holding unit and a current limiting feedback device, which has a response granularity of microseconds and an adaptive voltage control capability. During the injection process, those path levels that exhibit "non-linear growth" in the time stack are preferentially selected for gate installation. Non-linear growth refers to the propagation delay increase between two consecutive stack layers, which is more than twice the delay value of the previous layer, accompanied by a sudden drop in signal energy or a phase disturbance peak. After the injection of the hysteresis gate, the signal forwarding behavior of the node is time-limited, and the time stamp data fed back by its predecessor node needs to be subjected to accuracy judgment, stability detection and transmission legality confirmation before triggering the next level of data transmission action. This delay mechanism essentially forms a loop suppression buffer, which can form a short "freeze" state in the propagation path when the upstream node has loop feedback, effectively blocking the return path of the error data and preventing continuous oscillation of the network topology. All gate activation and release processes can be dynamically determined by time stamp information without additional human intervention, achieving automatic closed-loop control.

[0035] To verify whether the deployed dual-mirror time stamp and hysteresis gate have actual oscillation suppression effect, multiple rounds of transmission tests and path disturbance response tests are conducted on the shadow routing trajectory. The test data is randomly injected into the path source node by the vibration simulator, and the data packets are delivered in a bidirectional cross manner. After propagating forward to the destination, they are fed back to the source in reverse direction to verify the controllability of time stamp synchronization and gate delay response. During the test, the in-stack time, out-stack time, hysteresis gate delay time, data forwarding success rate and retransmission times of each hop node are recorded and compared with the control path without hysteresis gate. If in three consecutive tests, any node in the path does not appear reverse path misbuild, the number of communication cycle oscillations is less than once, the average propagation delay is stable within ±0.5 seconds, and the data complete transmission rate is not less than 98%, it is determined that the path has reached the topology stability control goal. Finally, the structure binds time and path behavior, builds a dynamic adjustable, response lag controllable and path closed loop breakable time anti-loop structure, providing a highly controllable time skeleton support for the next stage of topology energy constraint modeling and delay peak shifting forwarding mechanism.

[0036] S5, generating a time-varying virtual impedance route based on the time stack, building an energy landscape structure with a topology barrier, outputting a node forwarding sequence with a segmented peak shifting characteristic, thereby driving the wireless ad hoc network into a stable convergence state segment by segment; After completing the time stack construction, hysteresis gate deployment and path oscillation control, a dynamic forwarding mechanism with time sequence adjustment and topology guidance capability needs to be built to realize the stable state transition of the whole network communication. This is achieved by the following steps: The continuous communication records of each hop node in the time stack are extracted, the transmission and reception time intervals of each node at different time periods are statistically analyzed, and the average propagation delay of each node in the current communication period is calculated. The time delay value of each node is taken as the time sequence impedance quantitative index of the current path segment, and a node-time impedance table is formed. To improve the time domain controllability of path behavior, sliding section analysis is performed on the impedance data, and a time window of 5 seconds is taken. The impedance change curve of all nodes in the window is taken to construct a multi-cycle time-varying virtual impedance trajectory. The gradient value of the trajectory on the time axis reflects the state change of the communication load of the path with time evolution. If the impedance change amplitude of a node exceeds the set stable threshold (such as the growth rate of the previous and subsequent periods is greater than 25%), it is marked as a fluctuation point, which is an important reference node for subsequent adjustment control. The above steps realize the extraction of the dynamic change trend of the node time delay on the basis of the original time stack, and take it as the parameter starting point for building a stable path.

[0037] Based on the extracted time-varying virtual impedance trajectory, an energy landscape structure with spatial continuity and time fluctuation perception ability is constructed for the entire path segment. The construction method is as follows: taking time as the horizontal axis and the spatial position of the node as the vertical axis, the impedance value of each node is mapped to the three-dimensional height to form an impedance topographic map of the path segment. To enhance the physical interpretation ability of the topographic map, the historical energy distribution parameter of the node is introduced as a vertical direction correction factor, that is, the signal energy density change of the unit node in the past 20 seconds in the path is superimposed to the impedance height in the form of equal amplitude correction to generate a dynamic topological barrier map. In this map, the nodes where the impedance suddenly rises and the energy suddenly drops are regarded as barrier nodes, and their communication time sequence and energy conduction direction need to be delayed and intervened. At the same time, the energy low-lying area, that is, the node area with stable impedance and saturated signal strength, is identified as the activation source point of the preferential forwarding path. Through this map, the impedance highlands, energy channels and boundary transfer positions in the space-time dimension can be obtained, providing accurate reference for the next step of path peak shifting.

[0038] The "low impedance region" is defined by the time impedance value calculated from the average propagation delay of each node in the time stack. The lower the time impedance value of a node, the smaller the propagation delay of the node in the current communication period, the higher the signal transmission fluency, and the more stable the historical energy density and the weaker the coupling disturbance, which can be used as a preferential forwarding path. Therefore, the time impedance value of the node is mapped to its spatial position to form an impedance topographic map. In this topographic map, the nodes with impedance values in the preset low value interval are regarded as low impedance nodes.

[0039] The specific calibration method is as follows: the time impedance distribution of all nodes on the entire path is counted, the mean μ and the standard deviation σ are calculated, and the low impedance region is defined as μ kσ (k is an empirical coefficient, the value range can be 0.5 to 1.0) as a low impedance threshold. When the time impedance of the node is lower than the threshold, it is determined as a low impedance node; further, the low impedance nodes adjacent in space position are aggregated, that is, a low impedance region is formed. The region represents a path segment with high forwarding stability and low propagation delay in the topology structure, and is a data transmission segment to be activated preferentially in the subsequent peak-shifting scheduling plan.

[0040] According to the topology barrier map, the whole path is divided into multiple continuous but time behavior significantly different segments, and a segmented peak-shifting forwarding scheduling structure is constructed. The specific division standard is: if the time impedance difference between three consecutive nodes is not more than 10%, and the energy density is maintained in a set interval (such as-65dBm to-55dBm), it is divided into a stable sub-section; if the impedance of any node suddenly rises or the energy suddenly drops more than the threshold, it is the boundary of the section. Each stable sub-section is sorted according to the node impedance value, and the forwarding priority is assigned. The smaller the impedance value, the closer to the start time of the section, and the data forwarding is started first. The node forwarding trigger time interval is set to 1.2 times the maximum propagation delay of the section, in order to eliminate the data stacking effect and avoid loop path competition conflict. The forwarding scheduling table of all sub-sections is integrated into a path-level forwarding sequence plan, different sections are started in each time window, ensuring that the nodes are parallel in space, the hops are staggered in time, and the network is not staggered in structure, forming a time sequence scheduling network with peak-shifting trigger, segmented activation and predictable forwarding.

[0041] After the node segmented forwarding plan is completed, in order to realize the segment-by-segment convergence mechanism, a convergence verification window is set for each sub-section to observe whether the forwarding is completed successfully according to the predetermined time window. The convergence window starts from the start time of the first node and ends at the time period when the last node receives the data and feeds back the confirmation. If any node in the path fails to complete the data relay or feedback in the window, a synchronization correction signal is sent to the previous node, instructing it to lower the forwarding frequency or prolong the waiting period, and recalibrating the local timestamp and time stack reference. The sections that converge successfully are marked as stable path segments and no longer participate in scheduling adjustment; the sections that fail to converge for more than two rounds are marked as jitter path segments and are evaluated for secondary path replacement candidates. The whole process continues until all path segments are marked as stable state, completing the global coverage of the distributed forwarding plan. On this basis, all routing nodes in the network will start according to the segment, forward in staggered time, and verify segment by segment to enter the cooperative convergence state, completely getting rid of the problems of frequent loop reconstruction, path oscillation and data loss of the original ad hoc network under topology disturbance.

[0042] S6, after the network enters a stable convergence state, a time reversal phase traction operation is performed, inverse phase micro-pulses are injected to drive the programmable metasurface antenna array, and the shadow energy storage structure absorbs residual crosstalk energy, completing the broken ring control of the coherent topology and forming a sustainable closed-loop dynamic regulation mechanism; After completing the node forwarding peak-shifting control and making the network converge gradually, there may still be implicit closed-loop paths formed due to residual early interference. To completely eliminate such structures, the following steps are implemented: Based on the stable running path sequence, the phase residual records, round-trip delay change trajectories, and abnormal communication feedback information in the historical transmission data between nodes are retrieved to identify path segments that may have signal loopback trends. In this stage, the positions where asymmetric phase inversion (deviating from the uplink signal by 180 degrees) and average delay retraction (more than 20% lower than the previous period delay value) occur in the path nodes are analyzed as potential closed-loop hidden trouble points. Combined with the time stack data of the previous stage, time window analysis is performed on these areas to extract the time period when the signal appears to propagate in the reverse direction. Subsequently, multi-point time sequence cross sampling is performed on these nodes to capture the phase transition boundaries and power fluctuation peaks in the received signals, and compared with the communication time sequence of adjacent nodes to confirm whether there is an unauthorized reverse path. If the detected reverse phase echo signal has a periodic rebound phenomenon in three consecutive time windows, the node is listed in the list of reverse intervention targets and is used as a candidate injection node for subsequent phase traction.

[0043] Taking the identified reverse intervention node as the starting point, inverse phase micro-pulses are injected into its communication path as the driving signal for executing time reversal traction. Each injection pulse maintains the same frequency as the forward propagation data in signal parameters, with a waveform amplitude not less than 90% of the forward signal, a phase set to the standard reverse of the uplink signal (i.e., 180-degree out-of-phase), and a pulse width limited to no more than 200 nanoseconds, ensuring that the signal is short and has instantaneous energy release characteristics. The injection behavior of all micro-pulses strictly refers to the path start and end markers set in the time stack and is performed at the end of the communication period to avoid interference with the main path signal. After the pulse is emitted by the target node, it propagates in the reverse direction along the path, triggering phase response records in the passing nodes. If a node in the path detects reverse phase reconstruction (i.e., the detected waveform maintains phase synchronization with the inverse phase pulse), and this phenomenon persists for two communication periods, it is determined that the path has an unresolved loopback channel. This process can accurately identify closed-loop segments that are stable on the surface but have not yet been disconnected, avoiding potential topology risks due to path micro-coupling.

[0044] A programmable metasurface antenna array with active reverse energy manipulation capability is deployed for the detected closed loop path. The array is composed of multiple phased reflecting units, each with independent phase adjustment capability, which can continuously control the direction and phase state of the reflected wave within 0 to 360 degrees. The inversion intervention node is set as the center point, and no less than six reflecting unit points are deployed around it with a radius of 10 meters, forming a closed interference control structure. Each reflecting unit adjusts its reflection angle according to the time, phase and direction of the reverse phase signal it receives, so that its output waveform forms an interference field strength maximum value region, suppressing the signal's tendency to propagate in a loop on the path. At the same time, a shadow energy storage structure is deployed in the core area of the antenna array, which is composed of an energy absorption shell made of low-temperature ceramic material, a carbon nanotube charge trap layer and a multi-coil coupling buffer, which is used to specifically absorb high-frequency crosstalk pulses with a frequency of 8kHz to 20kHz, a phase misalignment of more than 90 degrees and a duration of less than 10 milliseconds, and convert them into non-propagating heat energy dissipation. Through the synergistic effect of antenna interference direction and shadow energy storage, the signals that originally propagate along the loop path are effectively dissipated or phase canceled in physical space, thereby blocking the possible formation of a loop closure path.

[0045] Combined with the inversion results and energy absorption efficiency evaluation, the entire network topology state is confirmed and dynamically updated for closed loop structure. After the inversion traction operation is completed, it is observed whether there is still a reverse phase response feedback in the path in the next three communication periods; if not, mark the path segment as "homodyne completed"; if there is still residual response, repeat the injection of pulses and adjust the antenna reflection parameters point by point until the echo is completely eliminated. Subsequently, according to the success state of each node broken loop, the forwarding priority is adjusted, and the broken loop node is moved to the preferred path of the main route relay, and the original connection direction with loop risk is shielded. Through this mechanism, the entire network finally forms a stable topology structure with only forward propagation as the only path, no secondary rebound, and controllable energy closed loop.

[0046] The application realizes accurate modeling of physical interference sources by constructing a time-frequency observation layer triggered by seismic events, real-time acquisition of electromagnetic interference characteristics, and extraction of crosstalk energy spectrum and time drift baseline; subsequently, using causal coherence decomposition and counterfactual playback chain technology, actively identifying and repairing distorted node links in the communication path, ensuring that the routing trajectory has stable convergence capability; further, by deploying double-mirror time scales and hysteresis gates to build a unified time stack structure, suppressing loop backflow and signal oscillation in topology reconstruction; on this basis, generating time-varying virtual impedance routing with segmented staggered peak characteristics, guiding the network to operate stably segment by segment; finally, with the help of programmable metasurface antenna arrays and shadow energy storage structures driven by inverse phase micro-pulse, active identification and physical ring breaking control of residual crosstalk paths are realized, and a coherent topology structure with long-term evolutionary adaptability and closed-loop control ability is constructed. This method improves the anti-interference, adaptability and topology stability of the wireless ad hoc network in the high-geological-disaster-chain area as a whole, and provides a solid technical guarantee for realizing high-frequency and high-precision disaster monitoring and early warning.

[0047] Certain exemplary embodiments of the present application have been described above by way of illustration, and it is to be understood that variations of the described embodiments can be made without departing from the spirit and scope of the present application. Therefore, the above figures and descriptions are illustrative in nature and are not to be construed as limiting the scope of the present application.

Claims

1. A method for deploying monitoring nodes for high-altitude geological disaster chains based on wireless ad hoc networks, characterized in that, Includes the following steps: S1, construct a time-frequency observation layer triggered by earthquakes in the high-level geological disaster chain area, deploy dual-path coupled probes and introduce phase calibration beacons, collect dynamic electromagnetic interference field information, and extract crosstalk energy spectrum and time drift baseline; S2, based on crosstalk energy spectrum and time drift baseline, performs causal coherence decomposition, identifies cross-channel coupling kernels and analyzes energy propagation direction, and identifies the set of sensitive nodes that cause routing anomalies; S3 constructs a counterfactual replay chain based on the set of sensitive nodes, replaces the signal distortion segments with shadow links, reconstructs a convergent routing trajectory, and restores the time series consistency between nodes; S4 deploys dual mirror time stamps around the routing trajectory to build a unified time stack and injects hysteresis gates into it to suppress oscillations caused by loop backflow and topology reconstruction; S5 generates time-varying virtual impedance routes based on the time stack, constructs a topological barrier energy landscape, outputs a node forwarding sequence with peak-shaving characteristics, and guides the network into a stable convergence state. S6, after the network enters a stable convergence state, performs time-reversal phase traction, drives the programmable metasurface antenna array by injecting inverse phase micropulses, and absorbs residual crosstalk energy in combination with the shadow energy storage structure.

2. The method for deploying high-altitude geological disaster chain monitoring nodes based on a wireless ad hoc network according to claim 1, characterized in that, Step S1 includes: Within the high-altitude geological hazard chain area, select typical areas with risks of landslides, collapses, or debris flows, and determine the layout boundary by combining geological survey results and topographic modeling data; Dual-path coupling probes are deployed at equal intervals within the deployment area. Each coupling probe collects disturbance electrical signals from the ground and the ground surface, and connects to a high-sensitivity sampling terminal via an independent wire. A set of phase calibration beacons is installed around each probe. After the earthquake is triggered, the beacons actively emit reference pulses with constant amplitude and frequency and consistent phase, and synchronously inject them into the adjacent probes to form a phase reference. After the earthquake disturbance occurs, the original disturbance waveforms of each channel are collected, and the amplitude change, phase shift and main peak energy density are extracted to generate crosstalk energy spectrum and time drift baseline.

3. The method for deploying high-altitude geological disaster chain monitoring nodes based on a wireless ad hoc network according to claim 1, characterized in that, Step S2 includes: Based on the crosstalk energy spectrum and time drift baseline, an interference time chain is constructed and a multi-node influence path map is generated; Based on the path map, the propagation levels are divided, frequency domain features are extracted, and node pairs with stable phase delay and high coherence amplitude are identified to construct a cross-channel coupled kernel database. A propagation path network is constructed based on the coupled kernel database and an energy propagation map is generated. An energy vector field is formed by combining the three-dimensional geographic coordinates. In the energy vector field, path aggregation points and disturbance abrupt change points are identified. The path oscillation response of nodes is verified by the disturbance simulation playback test method, and finally the set of sensitive nodes is determined.

4. The method for deploying high-altitude geological disaster chain monitoring nodes based on a wireless ad hoc network according to claim 1, characterized in that, Step S3 includes: Based on the set of sensitive nodes and the list of distorted channel segments, abnormal transmission paths are identified, and corresponding shadow links are constructed to replace the original path segments. The shadow link path is determined by hop count control, spatial boundary constraints, signal quality assessment, and return integrity rate screening. Node timing reconstruction is performed with the time drift baseline as a reference, and time alignment and error calibration are performed on all nodes in the shadow link. The shadow link is integrated into the main path architecture through path integration operations, and signal forwarding suppression mechanisms and routing table redirection control are applied.

5. The method for deploying high-altitude geological disaster chain monitoring nodes based on a wireless ad hoc network according to claim 4, characterized in that, The selection of shadow link paths must meet the following requirements: the number of hops should not exceed one and a half times the number of hops in the original path; the path space range should be limited to within fifty meters of the original path; the signal-to-noise ratio of all relay nodes should be higher than twenty-five decibels; and the signal return integrity rate should not be lower than 95%.

6. The method for deploying high-altitude geological disaster chain monitoring nodes based on a wireless ad hoc network according to claim 1, characterized in that, Step S4 includes: Dual mirror time stamp devices are deployed at both ends of the reconstructed route trajectory starting node and receiving node to complete time synchronization and direction verification between path jump points; Collect time-stamped data of all path nodes, construct a time stack structure, and calibrate the propagation delay and interference sensitivity coefficient to generate a routing time series matrix; Identify path segments in the time stack that exhibit non-linear growth characteristics, and inject hysteresis gates with adjustable response delays to form a forwarding buffer and path freezing mechanism. Multiple rounds of path transmission tests were conducted to verify the effects of time-scale synchronization, gate delay control, and path topology oscillation suppression.

7. The method for deploying high-altitude geological disaster chain monitoring nodes based on a wireless ad hoc network according to claim 6, characterized in that, The injection position of the hysteresis gate is limited to the position in the time stack where the propagation delay increase exceeds twice that of the previous level and is accompanied by a sudden drop in signal energy or a peak of phase disturbance, so as to ensure that the path oscillation source is precisely constrained.

8. The method for deploying high-altitude geological disaster chain monitoring nodes based on a wireless ad hoc network according to claim 1, characterized in that, Step S5 includes: Extract the average propagation delay of each path node in the time stack, generate a node time impedance lookup table, and construct a multi-period time-varying virtual impedance trajectory. By superimposing the impedance trajectories of each node with the historical energy distribution, a topological barrier map with spatial continuity and temporal fluctuation perception is formed, and barrier nodes and low impedance regions are identified. Based on impedance differences, the path segments are divided, a segmented staggered forwarding scheduling plan is formulated, and node forwarding is triggered sequentially according to impedance priority to generate a path-level forwarding sequence plan. Set a sub-segment convergence verification window, evaluate whether data transmission is completed according to schedule segment by segment, and adjust the forwarding frequency and timestamp based on the feedback results.

9. A method for deploying high-altitude geological disaster chain monitoring nodes based on a wireless ad hoc network according to claim 8, characterized in that, In the segmented peak-shifting forwarding scheduling plan, the node forwarding trigger time interval of each sub-segment is set to a preset multiple of the maximum propagation delay within that segment to avoid forwarding overlap between nodes. When any sub-segment fails to meet the convergence verification condition for two consecutive rounds, its path is marked as a jitter path segment, and a backup path replacement mechanism is activated to maintain the continuous and stable convergence of the overall network.

10. A method for deploying high-altitude geological disaster chain monitoring nodes based on a wireless ad hoc network according to claim 1, characterized in that, Step S6 includes: Identify nodes in the communication path that exhibit asymmetric phase reversal and delay retraction, extract the backpropagation time period, and establish a list of inversion intervention targets; Injecting inverse phase micropulse signals into the intervention target node and propagating them backward in time stack order at the end of the path to detect whether there is a persistent phase reconstruction response in the path; A programmable metasurface antenna array is deployed around the detection node, and the loop path is blocked in physical space through the synergistic effect of the phased reflector unit and the shadow energy storage structure. Adjust forwarding priorities based on the traction results and block risky directions to ensure that the entire network topology uses positive propagation as the only path, forming a dynamic control structure of closed-loop and closed-loop operation.

Images