An intelligent scheduling method for multi-parameter observation tasks of offshore buoy cluster

Abstract

The application discloses a kind of multi-parameter observation task intelligent scheduling methods for offshore buoy cluster, it is related to marine monitoring network cooperative scheduling technical field, for solving the data missing measurement and energy exhaustion problem caused by traditional buoy passive response, first, combine environmental parameter mutation gradient and power field vector to deduce abnormal event space diffusion path, generate with time label's propagation evolution vector and preheating delay sequence;Subsequently, through spatial intersection calculation accurate screening affected dynamic cooperative node set, combine node real-time available energy to construct resource constraint matrix;Fusion node historical observation coincidence degree, time delay compensation matching coefficient and resource constraint matrix cross operation generate task bidding benchmark sequence;Finally, rely on distributed network execution node bidding, to winning bidder node issue wake-up instruction containing start-up advance, make unwinning bidder node reduce frequency hibernate, realized the optimal dynamic balance of cluster observation utility and global energy consumption.

Description

Technical Field

[0001] This invention relates to the field of collaborative scheduling technology for marine monitoring networks, specifically to an intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters. Background Technology

[0002] Nearshore waters are active zones for marine ecological evolution and natural meteorological events, making continuous and comprehensive environmental parameter observation of these areas of significant application value. Nearshore buoy clusters, as core observation nodes deployed on the sea surface, are equipped with various types of environmental and hydrological sensors, enabling them to acquire wind fields, ocean currents, waves, and various water quality parameters around the clock. With the increasing demand for multidimensional ocean data, buoy cluster observation networks are gradually showing a trend towards denser nodes and more diversified sensors, placing higher demands on the synchronous acquisition and spatial coordination capabilities of multi-source data.

[0003] Existing buoy node task scheduling schemes typically employ static timed polling or passive response mechanisms triggered by single-node threshold exceedances. These methods suffer from substantial physical execution lag. Firstly, marine anomalies exhibit rapid spatial diffusion and propagation; isolated triggering of high-frequency sampling by a single node can lead to insufficient response from neighboring nodes, resulting in missed detection of crucial observational data during the early stages of pollution zones or environmental abrupt changes. Secondly, high-precision ocean quality or acoustic sensors require a lengthy hardware warm-up period before activation. Existing passive triggering scheduling fails to align the sensor hardware warm-up time with the spatial propagation time of the anomaly, causing the core area of ​​the anomaly to deviate from the designated sea area by the time the equipment is warmed up. Furthermore, indiscriminately waking up nearby nodes can cause network congestion and lead to blind concurrent operation of high-power sensors, resulting in premature depletion of the buoy node's available power. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters, thus solving the problems mentioned in the background.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters, comprising the following steps: S1. Collecting basic environmental parameters of the nearshore buoy cluster in continuous time series, extracting upstream nodes whose rate of change of basic environmental parameters exceeds a preset gradient threshold within adjacent sampling periods; obtaining environmental dynamic field vector data corresponding to the upstream nodes, calculating the spatial diffusion path and propagation speed in combination with the parameter change rate, generating an event propagation evolution vector with time evolution labels, and extracting the preheating time and power-on energy consumption of the corresponding sensors according to the parameter type corresponding to the parameter change rate to generate a multi-parameter preheating delay sequence; S2. Performing intersection calculation between the three-dimensional spatial coordinates of the nearshore buoy cluster and the event propagation evolution vector, extracting downstream nodes whose coordinates fall within the coverage area of ​​the event propagation evolution vector to form a dynamic collaborative node set; obtaining the real-time available energy status values ​​of the nodes in the dynamic collaborative node set, and combining the multi-parameter preheating delay sequence... S3. Extract the observation overlap of nodes within the dynamic collaborative node set within a preset historical time window and convert it into basic revenue weights. Extract the time evolution label from the event propagation evolution vector and perform a difference operation with the sensor warm-up time in the multi-parameter warm-up delay sequence to generate a delay compensation matching coefficient. Perform a weighted cross operation on the basic revenue weights, delay compensation matching coefficients, and node resource constraint matrix to generate a task bidding benchmark sequence for each node in the dynamic collaborative node set. S4. The upstream node broadcasts the observation task package containing the event propagation evolution vector to the dynamic collaborative node set. The nodes in the dynamic collaborative node set parse the observation task package and extract the values ​​in the task bidding benchmark sequence as bidding parameters for submission. Select the nodes whose bidding parameters meet the preset allocation rules as the winning nodes, issue a sensor wake-up command containing the power-on advance to the winning nodes, and issue a frequency reduction sleep command to the unsuccessful nodes.

[0006] Furthermore, the specific process for collecting continuous time series of basic environmental parameters of the nearshore buoy cluster and extracting upstream nodes whose rate of change of basic environmental parameters exceeds a preset gradient threshold within adjacent sampling periods is as follows: Retrieve low-frequency cruise observation data sequences from the basic environmental sensors of the nearshore buoy cluster within a preset time sliding window; extract the difference between the low-frequency cruise observation data sequences in adjacent sampling periods; combine this with the sampling time interval value to calculate the ratio, generating a basic environmental parameter change rate sequence; compare the basic environmental parameter change rate sequence with the preset gradient threshold; mark nodes in the basic environmental parameter change rate sequence that exceed the preset gradient threshold as upstream nodes; and extract the trigger parameter types corresponding to exceeding the preset gradient threshold.

[0007] Further, the following process is employed: First, the environmental dynamic field vector data corresponding to the upstream node is acquired. Second, the spatial diffusion path and propagation speed are calculated by combining the parameter change rate. Third, an event propagation evolution vector with time evolution labels is generated. Finally, based on the parameter type corresponding to the parameter change rate, the preheating time and power-on energy consumption of the corresponding sensor are extracted to generate a multi-parameter preheating delay sequence. The specific steps are as follows: First, the real-time surface current velocity vector and sea surface wind field vector of the upstream node are read, and vector synthesis is performed to generate environmental dynamic field vector data. Second, the diffusion influence coefficient corresponding to the trigger parameter type is extracted. Third, the environmental dynamic field vector data and the diffusion influence coefficient are multiplied to obtain the spatial diffusion path and propagation speed. Fourth, the expected arrival time is marked along the spatial diffusion path at equidistant spatial intervals to generate an event propagation evolution vector with time evolution labels. Finally, the target high-power sensor category associated with the trigger parameter type is queried in the preset sensor hardware mapping table. Fifth, the sensor preheating time and power-on energy consumption of the target high-power sensor category are extracted and combined to generate a multi-parameter preheating delay sequence.

[0008] Furthermore, the specific process of calculating the intersection of the three-dimensional spatial coordinates of the nearshore buoy cluster and the event propagation evolution vector, and extracting downstream nodes whose coordinates fall within the coverage of the event propagation evolution vector to form a dynamic collaborative node set is as follows: Extract the three-dimensional spatial coordinates of the upstream nodes as the reference vertex, and expand them to both sides along the spatial diffusion path of the event propagation evolution vector with a preset diffusion angle to generate a fan-shaped event propagation spatial envelope; Project the three-dimensional spatial coordinates of the nearshore buoy cluster onto a two-dimensional sea-level coordinate system to generate a sequence of two-dimensional projected coordinates of the nodes; Calculate the topological inclusion relationship between the sequence of two-dimensional projected coordinates of the nodes and the event propagation spatial envelope, and extract the node combinations whose two-dimensional projected coordinates are located inside the event propagation spatial envelope to generate a dynamic collaborative node set.

[0009] Furthermore, the specific process of obtaining the real-time available energy status values ​​of nodes within the dynamic collaborative node set and constructing the node resource constraint matrix by combining them with the sensor power-on energy consumption in the multi-parameter preheating delay sequence is as follows: Read the current battery state of charge value and real-time solar charging power value of the nodes within the dynamic collaborative node set, remove the node's preset basic sleep power consumption value, calculate and generate the real-time available energy status value of the nodes, and retrieve the sensor power-on energy consumption recorded in the multi-parameter preheating delay sequence; extract the node identifier within the dynamic collaborative node set as the row dimension element, extract the real-time available energy status value and sensor power-on energy consumption as the column dimension elements, and concatenate the data sets to output the node resource constraint matrix.

[0010] Furthermore, the specific process of extracting the observation overlap of nodes within the dynamic collaborative node set within a preset historical time window and converting it into basic revenue weights, extracting the time evolution label from the event propagation evolution vector, and performing a difference operation with the sensor preheating time in the multi-parameter preheating delay sequence to generate the delay compensation matching coefficient is as follows: Read the historical multi-parameter time series data of nodes within the dynamic collaborative node set within the preset historical time window, calculate the data covariance value of the historical multi-parameter time series data between nodes, obtain the physical spatial distance value between nodes, and generate the observation overlap value by solving the quotient based on the data covariance value and the physical spatial distance value; calculate the difference between the preset extreme value constant and the observation overlap value to generate the basic revenue weights; read the time evolution label recorded in the event propagation evolution vector, retrieve the corresponding sensor preheating time recorded in the multi-parameter preheating delay sequence, calculate the difference between the time evolution label and the sensor preheating time, and generate the delay compensation matching coefficient with an absolute timestamp format.

[0011] Furthermore, the specific process of generating the task bidding benchmark sequence for each node in the dynamic collaborative node set by performing weighted cross-operations on the basic revenue weight, latency compensation matching coefficient, and node resource constraint matrix is ​​as follows: Read the real-time available energy status value and sensor power-on energy consumption recorded in the node resource constraint matrix, calculate the ratio of the real-time available energy status value to the sensor power-on energy consumption, and generate the node energy availability coefficient; extract the revenue allocation ratio constant, latency allocation ratio constant, and energy allocation ratio constant from the preset configuration table, and perform weighted correlation calculations on the basic revenue weight, latency compensation matching coefficient, and node energy availability coefficient with the corresponding dimension's revenue allocation ratio constant, latency allocation ratio constant, and energy allocation ratio constant, respectively, and superimpose the output components of each weighted correlation calculation to generate the node comprehensive utility score; extract the node comprehensive utility score of each node in the dynamic collaborative node set, map and arrange them according to the node communication hardware physical address, and generate the task bidding benchmark sequence.

[0012] Furthermore, the upstream node broadcasts an observation task packet containing the event propagation evolution vector to the dynamic collaborative node set. The nodes within the dynamic collaborative node set parse the observation task packet and extract the values ​​from the task bidding benchmark sequence as bidding parameters for submission. The specific process is as follows: the event propagation evolution vector is encapsulated into the payload segment of a preset wireless communication data frame to generate an observation task packet. The observation task packet is then broadcast to the dynamic collaborative node set via the near-shore buoy cluster wireless self-organizing network channel. The nodes within the dynamic collaborative node set receive the observation task packet, perform communication protocol unpacking operations, extract the event propagation evolution vector obtained from the unpacking, and read the value in the task bidding benchmark sequence that matches their own communication hardware physical address as the node's comprehensive utility score. The node's comprehensive utility score is then encapsulated as a bidding parameter into a bid confirmation data packet and submitted to the upstream node.

[0013] Furthermore, the specific process of selecting nodes whose bidding parameters conform to the preset allocation rules as winning nodes, issuing sensor wake-up commands containing power-on advance values ​​to winning nodes, and issuing frequency reduction sleep commands to unsuccessful nodes is as follows: The upstream node receives the bid confirmation data packet, extracts the bidding parameters from the bid confirmation data packet, performs a descending sorting operation to generate a bid ranking list; compares the bid ranking list with the preset utility cutoff threshold defined in the preset allocation rules, selects nodes whose bidding parameters exceed the preset utility cutoff threshold as winning nodes, and selects nodes whose bidding parameters are lower than the preset utility cutoff threshold as unsuccessful nodes; extracts the delay compensation matching coefficient associated with the winning node and converts it into an absolute time countdown value as the power-on advance value, generates a sensor wake-up command containing the power-on advance value and a high-frequency sampling period value and sends it to the corresponding winning node; generates a frequency reduction sleep command containing an extremely low-frequency sleep sampling period value and sends it to the corresponding unsuccessful node.

[0014] The present invention has the following beneficial effects: (1) A multi-parameter observation task intelligent scheduling method for nearshore buoy clusters, by collecting upstream node data where the rate of change of basic environmental parameters exceeds a preset threshold, and combining environmental dynamic field vector to prospectively deduce event propagation evolution vector, changes the traditional passive waiting scheduling mode of event passing through, and realizes dynamic tracking of the diffusion path of abnormal environmental events; at the same time, based on this, the downstream nodes affected by the three-dimensional spatial intersection calculation are accurately delineated to form a dynamic collaborative node set, and a resource constraint matrix is ​​constructed by combining the required sensor power-on energy consumption and the available energy status of the nodes, so as to accurately focus the original blind network-wide response scheduling scope on the actual physical evolution and impact surface of the event, and avoid the invalid activation and energy waste of buoy nodes in unrelated sea areas from the source.

[0015] (2) A method for intelligent scheduling of multi-parameter observation tasks for nearshore buoy clusters. By extracting the observation overlap between nodes within a preset historical window and converting it into basic revenue weights, and creatively calculating the difference between the event arrival time tag and the preheating time of high-power sensors to generate a delay compensation matching coefficient, the method sends a sensor wake-up command containing a power-on advance amount to the winning node based on the distributed task bidding mechanism. This directly solves the technical problem of serious misalignment between the hardware preheating delay of high-power sensors and the movement speed of sudden events, ensuring that downstream nodes can complete equipment preheating and seamlessly switch to high-frequency observation state at the moment the event front arrives. At the same time, the method guides the redundant nodes that did not win the bid to enter the frequency reduction sleep state through the command, realizing a dynamic balance between maximizing the value of observation data and minimizing the overall power consumption of the equipment in the cluster network.

[0016] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0017] Figure 1 This is a flowchart of an intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters according to the present invention.

[0018] Figure 2 Flowchart for distributed task bidding and instruction issuance. Detailed Implementation

[0019] This application provides an intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters, which solves the problems of sensor preheating delay misalignment, missed detection of important feature data, and node power depletion caused by local blind concurrency due to the passive response mechanism of existing buoy networks when facing marine anomalies.

[0020] The overall concept of the solution in this application embodiment is as follows: First, by using upstream nodes whose basic environmental parameter change rate exceeds the threshold, combined with environmental dynamic field vectors, the spatial diffusion path and temporal evolution law of abnormal events are deduced, and the preheating time and energy consumption limit of the required sensors are mapped. Second, at the spatial level, the topology of downstream nodes falling within the coverage area of ​​the evolution vector is extracted, and their available energy is inventoried to construct a node resource constraint matrix. Then, the historical observation overlap of each node in the dynamic collaborative node set is quantified to clarify the benefit weight, and the event arrival countdown and sensor preheating time are physically aligned to generate a matching coefficient. Combined with the cross-operation of the resource constraint matrix, the task bidding benchmark sequence of each node is calculated. Finally, the upstream nodes broadcast observation task packages, and the downstream nodes submit bidding parameters according to the benchmark sequence. The system selects the winning node with the best comprehensive utility according to the preset allocation rules and issues a wake-up command with a power-on advance. The remaining unsuccessful nodes execute frequency reduction and hibernation, thereby completing the closed-loop intelligent control and optimal resource allocation of multi-parameter environmental collaborative observation.

[0021] Please see Figure 1This invention provides a technical solution: an intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters, comprising the following steps: S1. Collecting basic environmental parameters of the nearshore buoy cluster in continuous time series, extracting upstream nodes whose basic environmental parameter change rate exceeds a preset gradient threshold within adjacent sampling periods; obtaining environmental dynamic field vector data corresponding to the upstream nodes, calculating the spatial diffusion path and propagation speed in combination with the parameter change rate, generating an event propagation evolution vector with time evolution labels, and extracting the preheating time and power-on energy consumption of the corresponding sensors according to the parameter type corresponding to the parameter change rate to generate a multi-parameter preheating delay sequence; S2. Performing intersection calculation between the three-dimensional spatial coordinates of the nearshore buoy cluster and the event propagation evolution vector, extracting downstream nodes whose coordinates fall within the coverage area of ​​the event propagation evolution vector to form a dynamic collaborative node set; obtaining the real-time available energy status values ​​of the nodes in the dynamic collaborative node set, and combining the transmission in the multi-parameter preheating delay sequence... S3. Extract the observation overlap of nodes within the dynamic collaborative node set within a preset historical time window and convert it into basic revenue weights. Extract the time evolution label from the event propagation evolution vector and perform a difference operation with the sensor preheating time in the multi-parameter preheating delay sequence to generate a delay compensation matching coefficient. Perform a weighted cross operation on the basic revenue weights, delay compensation matching coefficients, and node resource constraint matrix to generate a task bidding benchmark sequence for each node in the dynamic collaborative node set. S4. The upstream node broadcasts the observation task package containing the event propagation evolution vector to the dynamic collaborative node set. Nodes within the dynamic collaborative node set parse the observation task package and extract the values ​​from the task bidding benchmark sequence as bidding parameters for submission. Select nodes whose bidding parameters meet the preset allocation rules as winning nodes. Send sensor wake-up instructions containing power-on advance to the winning nodes and frequency reduction sleep instructions to the unsuccessful nodes.

[0022] In this implementation plan, the main function of step S1 is to achieve early detection and spatiotemporal evolution trend prediction of marine anomalies. Environmental dynamic field vector data refers to the physical force parameters driving ocean water movement, including surface current velocity and direction with directional and magnitude attributes, as well as sea surface wind speed and direction. Event propagation evolution vector refers to a directional mathematical model describing the physical path and speed of marine anomalies spreading over time in a two-dimensional or three-dimensional spatial coordinate system. This step captures the abrupt gradient of basic environmental parameters of upstream nodes, combines this with the marine dynamic environment to calculate the actual physical trajectory of the anomaly, and pre-determines the hardware warm-up time and energy consumption limits required for the high-precision sensors that need to be linked. This process transforms the traditional passive over-limit alarm mechanism into a proactive prediction mechanism, providing precise spatiotemporal targets for subsequent resource allocation within the cluster. The main function of step S2 is to delineate the physical area affected by the anomaly and inventory the available hardware resources of the local cluster. The dynamic collaborative node set refers to the downstream buoy group combination temporarily selected in the three-dimensional spatial topology relationship based on the evolution trajectory deduced from the previous steps, which is located on the inevitable path of the future physical diffusion of the abnormal event. The node resource constraint matrix refers to the restrictive evaluation underlying table formed by multi-dimensional correspondence and data splicing of the current actual available power of the selected downstream buoys and the energy consumed by the high-power sensor's single power-up action. This step uses spatial intersection calculation to accurately focus the scheduling target on the actual physical impact surface of the sudden event, eliminating interference from irrelevant sea area nodes, and clarifying the energy base status of each physical node in the impact sea area, thus avoiding local communication congestion and unnecessary power waste caused by blind concurrent response of the entire cluster network. The main function of step S3 is to comprehensively and quantitatively evaluate the edge value of the observation data of the affected nodes and the hardware physical latency characteristics. The observation overlap refers to the spatial overlap and data redundancy of adjacent buoy nodes sampling the same marine environmental parameters within a historically set period. Higher redundancy indicates lower core value for conducting a new sampling. The time delay compensation matching coefficient is a compensation parameter generated by aligning the countdown of the physical front of an abnormal event reaching the target buoy's geographical location with the time required for the high-precision sensor on the buoy to warm up to a stable working state. The task bidding benchmark sequence is a comprehensive cost-effectiveness ranking score generated for each node to execute the current collaborative observation task, integrating data uniqueness benefits, time-physical matching degree, and energy margin. This step transforms multi-dimensional physical environmental parameters and hardware constraints into an intuitively calculable evaluation model, directly resolving the spatiotemporal misalignment technical pain point between the slow warm-up of high-power marine sensors and the rapid movement of sudden marine events. The main function of step S4 is to complete distributed task bidding and issue precise control commands to the underlying physical hardware devices based on the wireless ad hoc network.The power-on lead time refers to the absolute time countdown value calculated by the system based on the delay compensation matching coefficient. It is used to indicate how long before the target hardware device must forcibly start the power supply and preheating process before the arrival of the physical front of the abnormal event. The frequency reduction and hibernation command refers to the control information that guides non-critical buoy nodes to actively extend the sampling period to give up local communication bandwidth and save their own power. This step avoids the scheduling paralysis risk caused by the communication failure of the shore-based central server by using a decentralized collaborative execution method of upstream nodes broadcasting payload data and downstream nodes autonomously extracting utility scores for bidding. It also issues hardware control actions with precise power-on lead times to the winning nodes, which not only ensures that core high-value observation data is completely captured at the moment the event passes, but also effectively extends the overall fault-free endurance of the nearshore buoy cluster by putting the non-winning redundant nodes into hibernation.

[0023] Specifically, the process of collecting basic environmental parameters from a continuous time series of nearshore buoy clusters and extracting upstream nodes whose rate of change of basic environmental parameters exceeds a preset gradient threshold within adjacent sampling periods is as follows: Low-frequency cruise observation data sequences from the basic environmental sensors of the nearshore buoy cluster within a preset time sliding window are retrieved; the differences between the low-frequency cruise observation data sequences in adjacent sampling periods are extracted; and the ratio is calculated by combining the sampling time interval values ​​to generate a basic environmental parameter change rate sequence; the basic environmental parameter change rate sequence is compared with the preset gradient threshold, and nodes in the basic environmental parameter change rate sequence that exceed the preset gradient threshold are marked as upstream nodes; the trigger parameter types corresponding to exceeding the preset gradient threshold are extracted.

[0024] In this implementation scheme, the system initially operates in a routine energy-saving monitoring state. It acquires the fundamental physical parameters of nearshore buoy nodes within a set sliding time window—a low-frequency cruise observation data sequence—as the data source for determining the environmental background baseline. The difference between adjacent sampling periods in this sequence is extracted and combined with the time interval to calculate a ratio. This aims to filter out slow, natural environmental drift and accurately capture sudden step signals in marine environmental parameters. The specific calculation logic for this ratio is given by the formula... In the formula Rate of change of basic environmental parameters; : Buoy Node At the current time step Low-frequency cruise observation data values; : The low-frequency cruise observation data value of the same buoy node in the previous adjacent time step; The sampling time interval between adjacent sampling periods is represented by the numerical value. Subsequently, the system compares the generated rate of change of basic environmental parameters with a preset gradient threshold. If the threshold is exceeded, it is determined that an abnormal environmental mutation has occurred in the sea area where the node is located. The system then marks the node as the origin of the event, i.e., the upstream node, and extracts the specific physical or chemical parameter categories that caused the exceedance as trigger parameter types. The preset gradient threshold is determined by extracting historical background environmental observation data of the sea area in the same season within a preset period under conditions of no sudden event interference, calculating its statistical mean and standard deviation of fluctuation, and then using the formula... In the formula, Preset gradient threshold; Historical statistical average; : Standard deviation of historical data; This step involves setting a tolerance adjustment factor based on the periodic fluctuations of ocean currents in the area. By solving for the rate of change of the underlying data, this transforms the traditional passive polling mechanism into a data-driven active triggering mechanism, thus identifying the initial target for collaborative observation of the entire cluster.

[0025] Specifically, the process of acquiring environmental dynamic field vector data corresponding to upstream nodes, calculating spatial diffusion paths and propagation speeds by combining parameter change rates, generating event propagation evolution vectors with time evolution labels, and extracting the preheating time and power consumption of corresponding sensors based on the parameter types corresponding to the parameter change rates to generate multi-parameter preheating delay sequences is as follows: Read the real-time surface ocean current velocity vector and sea surface wind field vector of upstream nodes, and perform vector synthesis operations to generate environmental dynamic field vector data; extract the diffusion influence coefficients corresponding to the trigger parameter types, multiply the environmental dynamic field vector data and diffusion influence coefficients to obtain the spatial diffusion path and propagation speed, label the expected arrival time along the spatial diffusion path at equidistant spatial intervals, and generate event propagation evolution vectors with time evolution labels; query the target high-power sensor category associated with the trigger parameter type in the preset sensor hardware mapping table, extract the sensor preheating time and sensor power consumption of the target high-power sensor category, and combine them to generate multi-parameter preheating delay sequences.

[0026] In this implementation scheme, after identifying the upstream node and the type of anomalous parameters, the system reads two key meteorological and hydrological parameters—the real-time surface current velocity vector and the sea surface wind field vector—and performs physical-level spatial vector superposition to generate environmental dynamic field vector data. The aim is to construct the external driving force basis for the drift and evolution of anomalous matter or energy on the ocean surface. Next, the system introduces a diffusion influence coefficient strongly correlated with the triggering parameter type, and performs a product calculation on the environmental dynamic field vector data and the diffusion influence coefficient to deduce the actual physical spatial diffusion path and propagation speed of the anomalous event at the sea level. The calculation logic of this process is based on the formula... It means that in the formula : Event propagation evolution vector; : The diffusion effect coefficient calibrated for a specific trigger parameter type; Real-time surface ocean current velocity vector; : Wind-ocean current energy conversion coefficient; Real-time sea surface wind field vector. The diffusion influence coefficient is determined based on the physicochemical diffusion properties of specific environmental anomalies. For example, there are significant hydrodynamic differences between oil spill events and red tide algal blooms, requiring the extraction of corresponding constants from a pre-defined database of marine material hydrodynamic characteristics. Based on this, the system calculates the estimated arrival time of the anomaly front at each spatial coordinate node by dividing the calculated spatial diffusion path by the magnitude of the event propagation evolution vector at equidistant spatial intervals, thus forming an event propagation evolution vector with a time evolution label. Finally, the system queries a pre-defined sensor hardware mapping table to match the hardware startup hysteresis characteristics of high-power sensors, extracting the hardware warm-up time and single-start energy consumption of the corresponding target high-power sensors, and combining them into a multi-parameter warm-up delay sequence. This step deeply integrates and precisely quantifies the macroscopic evolution of ocean environmental flow field dynamics with the microscopic physical response delay characteristics of buoy hardware. This enables the system to accurately predict the future spatiotemporal distribution of abnormal events and grasp the energy consumption and time limit of the underlying equipment, providing an indispensable physical calculation basis for the subsequent accurate issuance of early wake-up commands in the affected sea area.

[0027] Specifically, the process of calculating the intersection of the three-dimensional spatial coordinates of the nearshore buoy cluster with the event propagation evolution vector, and extracting downstream nodes whose coordinates fall within the coverage of the event propagation evolution vector to form a dynamic collaborative node set is as follows: The three-dimensional spatial coordinates of the upstream nodes are extracted as reference vertices, and a preset diffusion angle is extended to both sides along the spatial diffusion path of the event propagation evolution vector to generate a fan-shaped event propagation spatial envelope; the three-dimensional spatial coordinates of the nearshore buoy cluster are projected onto a two-dimensional sea-level coordinate system to generate a sequence of two-dimensional projected coordinates of the nodes; the topological inclusion relationship between the sequence of two-dimensional projected coordinates of the nodes and the event propagation spatial envelope is calculated, and the nodes whose two-dimensional projected coordinates are located inside the event propagation spatial envelope are extracted to generate a dynamic collaborative node set.

[0028] In this implementation scheme, to accurately delineate the specific physical area affected by a marine emergency, the system first establishes the three-dimensional spatial coordinates of the upstream node as the baseline divergence vertex, which is equivalent to anchoring the zero-kilometer origin of the abnormal event in a vast sea area. Subsequently, the system expands to both sides of the spatial diffusion path of the event propagation evolution vector obtained in the previous step as the central axis, thus constructing a fan-shaped event propagation spatial envelope in three-dimensional space. The method for determining the preset diffusion angle is based on the statistical morphological characteristics of the diffusion profile of similar abnormal events in the historical sea area, combined with the current water flow turbulence intensity coefficient. Considering that nearshore environmental changes such as red tide proliferation or oil spill pollution mostly spread in two-dimensional surface advection along the sea surface, in order to reduce the computing power overhead of edge computing devices, the system vertically projects the original three-dimensional spatial coordinates of the nearshore buoy cluster to a two-dimensional sea-level coordinate system, thereby generating a dimensionality-reduced two-dimensional projected coordinate sequence of nodes. Based on this, the system performs topological inclusion relationship calculation between the node's two-dimensional projected coordinate sequence and the event propagation space envelope. This core calculation step transforms the physical geometric boundary of the sea area into a rigorous mathematical set selection rule to determine whether each target buoy is on the inevitable path of polluted or abnormal water bodies. The core spatial deflection angle characteristic in this topological inclusion relationship determination process is defined by the formula... The calculation shows that, in the formula The 2D projected coordinate sequence of a node The spatial deflection angle of each buoy node relative to the propagation axis of the abnormal event center; : Pointing from the reference divergence vertex to the first The planar physical distance vector of each buoy node; The normalized unit direction vector of the event propagation evolution vector. When the calculated spatial deflection angle is less than or equal to half of the preset diffusion angle, and the physical distance between the buoy and the reference divergence vertex is less than the maximum propagation range of the event within a specific time window, the system determines that the two-dimensional projected coordinates completely fall within the event propagation spatial envelope. Through the above steps, the system successfully extracts the node combinations that meet the conditions and generates a dynamic cooperative node set. This completely excludes safe sea area nodes outside the event propagation path, effectively preventing large-scale ineffective scheduling and waste of computing resources caused by blind wake-ups across the entire network.

[0029] Specifically, the process of obtaining the real-time available energy status values ​​of nodes within the dynamic collaborative node set and constructing the node resource constraint matrix by combining them with the sensor power-on energy consumption in the multi-parameter preheating delay sequence is as follows: Read the current battery state of charge value and real-time solar charging power value of the nodes within the dynamic collaborative node set, remove the node's preset basic sleep power consumption value, calculate and generate the real-time available energy status value of the nodes, and retrieve the sensor power-on energy consumption recorded in the multi-parameter preheating delay sequence; extract the node identifier within the dynamic collaborative node set as the row dimension element, extract the real-time available energy status value and sensor power-on energy consumption as the column dimension elements, and concatenate the data sets to output the node resource constraint matrix.

[0030] In this implementation plan, after identifying the dynamic collaborative node set located in high-risk sea areas, the system must assess and quantitatively reconstruct the hardware resource maintenance capabilities of these nodes under adverse sea conditions to ensure the physical executability of subsequent scheduling commands. The system first reads the current battery state of charge and real-time solar charging power values ​​fed back by the underlying hardware management modules of each node in the dynamic collaborative node set. This multi-source data acquisition is necessary because the buoy platform is deployed in the field for extended periods, and its energy supply is not a static constant but a variable that dynamically replenishes based on sunlight conditions. To obtain the net available energy that a node can use to perform emergency observation tasks, the system removes the node's preset basic dormancy power consumption value, which is essential for the buoy's basic survival. This preset basic dormancy power consumption value is determined by summing the static maintenance current of the main control circuit board, the sensor's reference standby current, and the periodic heartbeat transmission power consumption of the satellite communication module under standard test conditions. The core logic for calculating the node's actual energy reserves is based on the formula... The calculation shows that, in the formula : Dynamic collaborative node centralization Real-time available energy status values ​​for each buoy node; : Total design physical capacity constant of the battery pack mounted on the buoy node; No. The ratio of the current battery state of charge of each buoy node; :No. Real-time solar charging power values ​​fed back by the photovoltaic energy storage management components of each buoy node; No. The preset basic sleep power consumption value corresponding to each buoy node; This calculation determines the estimated total duration of high-frequency observation tasks required to address this marine anomaly. Using this calculation system, the system calculates the net remaining power available to each node after basic self-preservation. Subsequently, the system retrieves the pre-recorded high-precision sensor power-on energy consumption from the multi-parameter preheating delay sequence and uses the network communication media access control hardware identifier of each independent node in the dynamic collaborative node set as a row-dimensional element for spatial addressing mapping. Simultaneously, the calculated real-time available energy status value and the sensor power-on energy consumption to be consumed are extracted and used as column-dimensional elements to fill the corresponding mapping area. Finally, after rigorous data assembly and encapsulation, a node resource constraint matrix is ​​output. The successful construction of this matrix essentially establishes an energy income and expenditure ledger within the buoy physical computing platform. It intuitively presents the comparison between the remaining power of each standby node before power depletion and the cost required to execute a new task, laying a solid data foundation for subsequent selection of the optimal cost-effective node based on distributed algorithms.

[0031] Specifically, the process of extracting the observation overlap of nodes within a preset historical time window and converting it into basic revenue weights, extracting the time evolution label from the event propagation evolution vector, and performing a difference operation with the sensor preheating time in the multi-parameter preheating delay sequence to generate a delay compensation matching coefficient is as follows: Read the historical multi-parameter time series data of nodes within the preset historical time window, calculate the data covariance value of the historical multi-parameter time series data between nodes, obtain the physical spatial distance value between nodes, and generate the observation overlap value by solving the quotient based on the data covariance value and the physical spatial distance value; calculate the difference between the preset extreme value constant and the observation overlap value to generate basic revenue weights; read the time evolution label recorded in the event propagation evolution vector, retrieve the corresponding sensor preheating time recorded in the multi-parameter preheating delay sequence, calculate the difference between the time evolution label and the sensor preheating time, and generate a delay compensation matching coefficient with an absolute timestamp format.

[0032] In this implementation scheme, to eliminate redundant buoys that are spatially close and have highly similar observation data, it is necessary to evaluate the data edge value of each node. The system first reads the historical multi-parameter time series data of each node within a preset historical time window, calculates the data covariance of this historical multi-parameter time series data between nodes, and obtains the physical spatial distance between nodes. To measure the degree of joint sensing redundancy of adjacent nodes for the same ocean phenomenon, the system generates an observation overlap value by solving the quotient of the data covariance and physical spatial distance values. This value is then further transformed into a basic benefit weight for measuring the value of a node initiating an observation mission. The core calculation logic of this transformation process is expressed by the formula... Expression, in the formula : Dynamic collaborative node centralization The basic revenue weight of each node; Preset extreme value constant; The collection space of dynamic collaborative node sets; : The data covariance value between the x-th node and the y-th node in the historical multi-parameter time series data within the preset historical time window; :No. The node and the first The physical spatial distance values ​​between nodes are calculated. The preset extreme value constant is determined by extracting the upper limit of the statistical extreme value of the quotient of the maximum normalized covariance and the minimum physical distance from the historical sensor network data of the target sea area. This step, through mathematical spatial transformation, ensures that nodes with high historical data similarity and close proximity receive lower reward weights, thereby ensuring that computing power and electrical energy are tilted towards nodes that can provide unique observation data. Simultaneously, considering the physical startup delay of high-precision marine observation equipment, the system reads the time evolution tag recorded in the event propagation evolution vector of the preceding steps and retrieves the corresponding sensor preheating time recorded in the multi-parameter preheating delay sequence. By calculating the difference between the time evolution tag and the sensor preheating time, a delay compensation matching coefficient with an absolute timestamp format is generated. This step accurately calculates the absolute time node at which the equipment must be powered on at the latest, completely solving the physical misalignment defect caused by the slow preheating of high-power marine sensors leading to missed detection of the core front of sudden events.

[0033] Specifically, the process of generating the task bidding benchmark sequence for each node in the dynamic collaborative node set by performing weighted cross-operations on the basic revenue weight, latency compensation matching coefficient, and node resource constraint matrix is ​​as follows: Read the real-time available energy status value and sensor power-on energy consumption recorded in the node resource constraint matrix, calculate the ratio of the real-time available energy status value to the sensor power-on energy consumption, and generate the node energy availability coefficient; extract the revenue allocation ratio constant, latency allocation ratio constant, and energy allocation ratio constant from the preset configuration table, and perform weighted correlation calculations on the basic revenue weight, latency compensation matching coefficient, and node energy availability coefficient with the corresponding dimension's revenue allocation ratio constant, latency allocation ratio constant, and energy allocation ratio constant, respectively, and superimpose the output components of each weighted correlation calculation to generate the node comprehensive utility score; extract the node comprehensive utility score of each node in the dynamic collaborative node set, map and arrange them according to the node communication hardware physical address, and generate the task bidding benchmark sequence.

[0034] In this implementation scheme, to select the globally optimal solution for performing collaborative observation tasks from a limited pool of buoy node resources, the system must conduct a multi-objective optimization evaluation of the execution capabilities and data value of each node. The system first reads the real-time available energy status values ​​and sensor power-on energy consumption recorded in the node resource constraint matrix, and generates a node energy availability coefficient by calculating the ratio of the real-time available energy status values ​​to the sensor power-on energy consumption. This operation clarifies the hardware capabilities of each standby buoy to support several high-frequency power-on sampling sessions while maintaining its basic dormant survival status. Subsequently, the system extracts the revenue allocation ratio constant, latency allocation ratio constant, and energy allocation ratio constant from a preset configuration table, and performs weighted correlation calculations on the basic revenue weight, latency compensation matching coefficient, and node energy availability coefficient generated in the previous steps. The system then superimposes the output components of each weighted correlation calculation to generate a node comprehensive utility score that measures the overall scheduling priority of each node. This core multi-objective weighted cross-operation logic is expressed by the formula: In the formula, : Dynamic collaborative node centralization The overall utility score of each buoy node; :Profit distribution ratio constant; :No. The basic revenue weight of each buoy node; Delay allocation ratio constant; :No. The delay compensation matching coefficient for each buoy node; Energy distribution ratio constant; :No. The real-time available energy status values ​​of each buoy node in the resource constraint matrix; The first event required to trigger the current observation task. The sensor power-on energy consumption of the sensor-like system. The determination of the three weighted constants—revenue allocation ratio constant, delay allocation ratio constant, and energy allocation ratio constant—is based on the emergency response level of current marine emergencies, using the analytic hierarchy process (AHP) to calculate the eigenvectors of the relative importance matrix of different physical optimization objectives. Finally, the system extracts the comprehensive utility score of each node in the dynamic collaborative node set and maps and arranges them at the underlying data link layer according to the physical addresses of the node communication hardware, generating the final task bidding benchmark sequence for distributed bidding. This step perfectly integrates complex multi-dimensional marine environmental factors, hardware preheating time delays, and battery remaining energy constraints into an intuitive mathematical competition scoring system, enabling the collaborative scheduling of nearshore buoy clusters to break free from the rigid limitations of a single power threshold and possess intelligent attributes driven by multi-source heterogeneous data.

[0035] Please see Figure 2Specifically, the upstream node broadcasts an observation task packet containing the event propagation evolution vector to the dynamic collaborative node set. Nodes within the dynamic collaborative node set parse the observation task packet and extract values ​​from the task bidding benchmark sequence as bidding parameters for submission. The specific process is as follows: the event propagation evolution vector is encapsulated into the payload segment of a preset wireless communication data frame to generate an observation task packet, which is then broadcast to the dynamic collaborative node set via the near-shore buoy cluster wireless ad hoc network channel. Nodes within the dynamic collaborative node set receive the observation task packet, perform communication protocol unpacking operations, extract the unpacked event propagation evolution vector, and read the value matching their own communication hardware physical address from the task bidding benchmark sequence as their node comprehensive utility score. They then encapsulate this node comprehensive utility score as bidding parameters into a bid confirmation data packet and submit it to the upstream node.

[0036] In this implementation scheme, to achieve decentralized task allocation and coordinated operation of underlying devices, the system relies on the wireless self-organizing network channel of the near-shore buoy cluster for interactive communication. Upstream nodes act as temporary bidding control terminals at this stage, encapsulating the event propagation evolution vector accurately calculated in the preceding steps as core payload data into the payload segment of a preset wireless communication data frame. This generates a standardized observation task package, which is then broadcast to the dynamic collaborative node set. This broadcast mechanism completely eliminates the traditional ocean observation communication topology that heavily relies on shore-based central servers for unified scheduling, effectively avoiding the risk of the entire monitoring network being paralyzed due to the disconnection or high latency of offshore satellite communication links. When each downstream physical node in the dynamic collaborative node set receives the observation task package, the embedded microprocessor inside the node immediately performs a communication protocol unpacking operation, accurately extracting the event propagation evolution vector to understand the spatial dynamics of abnormal events. Subsequently, the underlying program of each node reads the task bidding benchmark sequence generated in the preceding calculations, and uses a verification mechanism matching its own communication hardware physical address to accurately extract the value specifically belonging to its own node as the node's comprehensive utility score. Ultimately, the downstream node encapsulates the node's overall utility score as a core bidding parameter into a bid confirmation data packet and submits it to the upstream node via the underlying media access control layer protocol. This step, through precise mapping of hardware addresses and a fully distributed communication bidding mechanism, enables each buoy in high-risk waters to independently assess its own combat value and report it, greatly improving the buoy swarm's responsiveness and network robustness in the face of sudden marine environmental events.

[0037] Specifically, the process of selecting nodes whose bidding parameters conform to the preset allocation rules as winning nodes, issuing sensor wake-up commands containing power-on advance values ​​to winning nodes, and issuing frequency reduction sleep commands to unsuccessful nodes is as follows: The upstream node receives the bid confirmation data packet, extracts the bidding parameters from the bid confirmation data packet, performs a descending sorting operation to generate a bid ranking list; compares the bid ranking list with the preset utility cutoff threshold defined in the preset allocation rules, selects nodes whose bidding parameters exceed the preset utility cutoff threshold as winning nodes, and selects nodes whose bidding parameters are lower than the preset utility cutoff threshold as unsuccessful nodes; extracts the delay compensation matching coefficient associated with the winning node and converts it into an absolute time countdown value as the power-on advance value, generates a sensor wake-up command containing the power-on advance value and a high-frequency sampling period value and sends it to the corresponding winning node; generates a frequency reduction sleep command containing an extremely low-frequency sleep sampling period value and sends it to the corresponding unsuccessful node.

[0038] In this implementation scheme, after collecting the bid confirmation data packets from each downstream node, the upstream node needs to make a final scheduling decision from a physical perspective of global resource optimization. The microprocessor of the upstream node extracts the bidding parameters from all bid confirmation data packets and performs a descending sort operation based on the quicksort algorithm to generate an intuitive bid ranking list. In order to select the most observable node combination within the limited communication bandwidth and computing power pool, the system compares the bid ranking list with the preset utility cutoff threshold defined in the preset allocation rules, and then selects nodes whose bid parameters exceed the preset utility cutoff threshold as winning nodes, while selecting nodes whose bid parameters are lower than the preset utility cutoff threshold as unsuccessful nodes. The preset utility cutoff threshold is determined by real-time monitoring of the available remaining physical bandwidth of the current wireless ad hoc network channel, dividing it by the standard uplink communication channel bandwidth required by a single node under high-frequency sampling conditions, to obtain the maximum number of concurrent nodes that the current network can theoretically accommodate simultaneously. Then, the bid parameter values ​​at the positions in the bid ranking list that are equal to this number constant are accurately extracted as the threshold. For the selected winning nodes, the system needs to convert their associated latency compensation matching coefficients into physical time control parameters that can be directly executed by the underlying hardware. The specific conversion and calculation logic of the core absolute time countdown value is given by the formula: In the formula :No. The absolute time countdown value corresponding to each winning node is the startup lead time; No. The absolute timestamp format value carried in the delay compensation matching coefficient of each winning node; The absolute reference time value currently synchronized by the nearshore buoy cluster system; The system calculates the time synchronization compensation error boundary value for the underlying wireless ad hoc network channel hardware. Based on the obtained power-on advance and high-frequency sampling period values, the system encapsulates and generates sensor wake-up commands and sends them to the corresponding winning nodes. This ensures that high-power, high-precision sensors can be powered on and preheated at a precise time before the arrival of abnormal water bodies, reaching their optimal working state for ultra-high-frequency data acquisition at the exact moment the pollution front passes. Simultaneously, the system generates a frequency-reducing sleep command containing ultra-low-frequency sleep sampling period values ​​and sends it to the corresponding non-winning nodes, forcing them to shut down unnecessary power-consuming hardware and significantly extending the sampling interval. This step ensures that the best resources for observation within the cluster are used effectively, guaranteeing the complete capture of core high-value feature data during spatiotemporal evolution, while significantly reducing ineffective energy consumption across the entire network through large-scale coordinated sleep.

[0039] In summary, this application has at least the following effects: A novel intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters is proposed. This method actively extrapolates the spatiotemporal propagation trajectory of anomalous events by integrating the gradient of abrupt changes in basic environmental parameters and the vector of the ocean dynamic field. It then uses three-dimensional spatial topological intersection calculations to accurately select affected downstream buoys to form a dynamic collaborative node set, thus avoiding localized network communication congestion and unnecessary power consumption caused by blind concurrent responses across the entire network from a physical source. Furthermore, it innovatively calculates the difference between the expected arrival time stamp of the event and the hardware physical warm-up time of high-power sensors to generate a delay compensation matching coefficient, completely resolving the issue of missed measurements caused by slow warm-up of high-precision oceanographic instruments. The system addresses the technological shortcomings of spatiotemporal misalignment at the core front of an event. Furthermore, it establishes a fully distributed task bidding mechanism based on the underlying wireless self-organizing network. This mechanism performs multi-objective weighted cross-operations on the data edge observation value of nodes, latency compensation matching degree, and the real-time available energy status of the underlying layer to generate a comprehensive utility score. This accurately guides the winning nodes to wake up the equipment in advance according to an absolute countdown to perform high-frequency sampling, and forces the redundant nodes that did not win the bid to enter a frequency-reduced deep sleep state. This achieves the optimal dynamic balance between the high-value collaborative observation data capture rate and the global fault-free survival cycle of the buoy cluster under the constraints of severely limited marine IoT communication bandwidth and node basic power.

[0040] 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 embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0041] This invention is described with reference to flowchart illustrations and / or block diagrams of systems, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0042] 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.

[0043] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0044] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0045] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for intelligent scheduling of multi-parameter observation tasks for nearshore buoy clusters, characterized in that, Includes the following steps: S1. Collect basic environmental parameters of the nearshore buoy cluster in continuous time series, extract upstream nodes whose basic environmental parameter change rate exceeds the preset gradient threshold in adjacent sampling periods; obtain environmental dynamic field vector data corresponding to the upstream nodes, calculate the spatial diffusion path and propagation speed in combination with parameter change rate, generate event propagation evolution vector with time evolution label, and extract the preheating time and power-on energy consumption of the corresponding sensor according to the parameter type corresponding to the parameter change rate to generate a multi-parameter preheating delay sequence; S2. The intersection of the three-dimensional spatial coordinates of the nearshore buoy cluster and the event propagation evolution vector is calculated, and downstream nodes whose coordinates fall within the coverage of the event propagation evolution vector are extracted to form a dynamic collaborative node set; that is, the three-dimensional spatial coordinates of the upstream nodes are extracted as the reference vertex, and the preset diffusion angle is expanded to both sides along the spatial diffusion path of the event propagation evolution vector to generate a fan-shaped event propagation spatial envelope. The three-dimensional spatial coordinates of the nearshore buoy cluster are projected onto a two-dimensional sea-level coordinate system to generate a sequence of two-dimensional projected coordinates of the nodes; Calculate the topological inclusion relationship between the two-dimensional projected coordinate sequence of the execution node and the envelope of the event propagation space, extract the node combination whose two-dimensional projected coordinates are located inside the envelope of the event propagation space to generate a dynamic cooperative node set; obtain the real-time available energy status value of the nodes in the dynamic cooperative node set, and construct the node resource constraint matrix by combining the sensor power-on energy consumption in the multi-parameter preheating delay sequence. S3. Extract the observation overlap of nodes in the dynamic collaborative node set within the preset historical time window and convert it into basic revenue weight. That is, calculate the difference between the preset extreme value constant and the observation overlap value to generate basic revenue weight. Extract the time evolution label in the event propagation evolution vector and perform a difference operation with the sensor preheating time in the multi-parameter preheating delay sequence to generate a delay compensation matching coefficient. The basic revenue weights, latency compensation matching coefficients, and node resource constraint matrices are weighted and cross-operated to generate the task bidding benchmark sequence for each node in the dynamic collaborative node set. The specific process of generating this sequence is as follows: Read the real-time available energy status value and sensor power-on energy consumption recorded in the node resource constraint matrix, calculate the ratio of the real-time available energy status value to the sensor power-on energy consumption, and generate the node energy availability coefficient. Extract the revenue distribution ratio constant, latency distribution ratio constant, and energy distribution ratio constant from the preset configuration table. Perform weighted correlation calculations between the basic revenue weight, latency compensation matching coefficient, and node energy availability coefficient and the corresponding dimension's revenue distribution ratio constant, latency distribution ratio constant, and energy distribution ratio constant. Superimpose the output components of each weighted correlation calculation to generate the node's comprehensive utility score. Extract the comprehensive utility score of each node in the dynamic collaborative node set, map and arrange them according to the physical address of the node communication hardware, and generate a task bidding benchmark sequence; S4. The upstream node broadcasts the observation task package containing the event propagation evolution vector to the dynamic collaborative node set. The nodes in the dynamic collaborative node set parse the observation task package, extract the values ​​in the task bidding benchmark sequence as bidding parameters and submit them. The nodes whose bidding parameters meet the preset allocation rules are selected as the winning nodes. The winning nodes are sent a sensor wake-up command containing the power-on advance amount, and the unsuccessful nodes are sent a frequency reduction and sleep command.

2. The intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters according to claim 1, characterized in that: The specific process for collecting basic environmental parameters from a continuous time series of nearshore buoy clusters and extracting upstream nodes whose rate of change of basic environmental parameters exceeds a preset gradient threshold within adjacent sampling periods is as follows: The low-frequency cruise observation data sequence of the nearshore buoy cluster basic environmental sensor within a preset time sliding window is retrieved, the difference between the low-frequency cruise observation data sequence in adjacent sampling periods is extracted, and the ratio is calculated by combining the sampling time interval value to generate a basic environmental parameter change rate sequence. By comparing the rate of change sequence of basic environmental parameters with the preset gradient threshold, nodes in the rate of change sequence of basic environmental parameters that exceed the preset gradient threshold are marked as upstream nodes, and the trigger parameter types corresponding to exceeding the preset gradient threshold are extracted.

3. The intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters according to claim 2, characterized in that: The specific process of obtaining environmental dynamic field vector data corresponding to upstream nodes, calculating spatial diffusion paths and propagation speeds by combining parameter change rates, generating event propagation evolution vectors with time evolution labels, and extracting the preheating time and power-on energy consumption of corresponding sensors based on the parameter type corresponding to the parameter change rate to generate a multi-parameter preheating delay sequence is as follows: Read the real-time surface ocean current velocity vector and sea surface wind field vector from the upstream node, and perform vector synthesis operation to generate environmental dynamic field vector data; Extract the diffusion influence coefficient corresponding to the trigger parameter type, multiply the environmental dynamic field vector data and the diffusion influence coefficient to obtain the spatial diffusion path and propagation speed, mark the expected arrival time along the spatial diffusion path at equal spatial intervals, and generate an event propagation evolution vector with time evolution labels; Query the target high-power sensor category associated with the trigger parameter type in the preset sensor hardware mapping table, extract the sensor warm-up time and sensor power-on power consumption of the target high-power sensor category, and combine them to generate a multi-parameter warm-up delay sequence.

4. The intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters according to claim 1, characterized in that: The specific process of obtaining the real-time available energy status values ​​of nodes within the dynamic collaborative node set, and combining them with the sensor power-on energy consumption in the multi-parameter preheating delay sequence to construct the node resource constraint matrix is ​​as follows: Read the current battery state of charge value and real-time solar charging power value of the nodes in the dynamic collaborative node set, strip the preset basic sleep power consumption value of the nodes, calculate the real-time available energy state value of the generated nodes, and retrieve the sensor power-on energy consumption recorded in the multi-parameter preheating delay sequence. Extract the node identifiers within the dynamic collaborative node set as row-dimensional elements, and extract the real-time available energy status values ​​and sensor power-on energy consumption as column-dimensional elements. Concatenate the data sets to output the node resource constraint matrix.

5. The intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters according to claim 1, characterized in that: The specific process of extracting the observation overlap of nodes within the dynamic collaborative node set within a preset historical time window and converting it into basic revenue weights, extracting the time evolution label from the event propagation evolution vector, and performing a difference operation with the sensor preheating time in the multi-parameter preheating delay sequence to generate the delay compensation matching coefficient is as follows: Read the historical multi-parameter time series data of nodes in the dynamic collaborative node set within a preset historical time window, calculate the data covariance value of the historical multi-parameter time series data between nodes, obtain the physical spatial distance value between nodes, and generate the observation overlap value by solving the quotient based on the data covariance value and the physical spatial distance value. The difference between the preset extreme constant and the observed overlap value is calculated to generate the basic benefit weight. The time evolution tag recorded in the event propagation evolution vector is read, the corresponding sensor warm-up time recorded in the multi-parameter warm-up delay sequence is retrieved, the difference between the time evolution tag and the sensor warm-up time is calculated, and the delay compensation matching coefficient with absolute timestamp format is generated.

6. The intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters according to claim 1, characterized in that: The upstream node broadcasts an observation task package containing the event propagation evolution vector to the dynamic collaborative node set. The nodes within the dynamic collaborative node set parse the observation task package and extract the values ​​from the task bidding benchmark sequence as bidding parameters for submission. The specific process is as follows: The event propagation evolution vector is encapsulated into the payload segment of a preset wireless communication data frame to generate an observation task package. The observation task package is then broadcast to the dynamic cooperative node set via the wireless self-organizing network channel of the near-shore buoy cluster. Nodes in the dynamic collaborative node set receive observation task packets, perform communication protocol unpacking operations, extract the event propagation evolution vector obtained from the unpacking, and read the value of matching their own communication hardware physical address in the task bidding benchmark sequence as the node's comprehensive utility score. The node's comprehensive utility score and the corresponding delay compensation matching coefficient are combined as bidding parameters, encapsulated into the bidding confirmation data packet, and submitted to the upstream node.

7. The intelligent scheduling method for multi-parameter observation tasks of nearshore buoy clusters according to claim 6, characterized in that: The specific process of selecting nodes whose bidding parameters meet the preset allocation rules as winning nodes, issuing sensor wake-up commands containing power-on advance amounts to winning nodes, and issuing frequency reduction and sleep commands to unsuccessful nodes is as follows: The upstream node receives the bid confirmation data packet, extracts the bid parameters from the bid confirmation data packet, performs a descending sort operation to generate a bid sorting list; Compare the bid ranking list with the preset utility cutoff threshold defined in the preset allocation rules, select nodes whose bid parameters exceed the preset utility cutoff threshold as winning nodes, and select nodes whose bid parameters are lower than the preset utility cutoff threshold as unsuccessful nodes. Extract the delay compensation matching coefficient associated with the winning node and convert it into an absolute time countdown value as the power-on advance. Generate a sensor wake-up command containing the power-on advance and a high-frequency sampling period value and send it to the corresponding winning node. Generate a frequency reduction sleep command containing the value of the extremely low frequency sleep sampling period and send it to the corresponding unsuccessful node.

Images