Water level monitoring device power consumption autonomous regulation method and system using context awareness

Abstract

The application relates to the technical field of Internet of Things device power consumption management, and particularly discloses a water level monitoring device power consumption self-regulation method and system adopting context situation awareness. First, a multi-level context situation awareness layer and model are constructed by combining local historical hydrological data with low-power access to multi-source external information; then, multi-dimensional operation data of the device are collected, a coupling relationship between the multi-dimensional operation data and the situation model is analyzed, power consumption value is evaluated to determine a regulation level; next, device modules are abstracted into energy consumption nodes, a power consumption topology is constructed, and a power consumption reduction scheme group is generated; subsequently, a digital twin is built, a scheme resilience index is simulated and evaluated, and the optimal scheme is selected; finally, the optimal scheme is deployed, the environment and the strategy effect are continuously monitored, and when a triggering condition is met, backtracking iteration is performed, so that precise and adaptive regulation is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power consumption management of Internet of Things devices, and particularly relates to a water level monitoring device power consumption autonomous regulation method and system using context awareness. BACKGROUND

[0002] Water level monitoring devices are key infrastructure in the fields of hydrological monitoring, flood warning, water resource management, etc., and are widely deployed in rivers, lakes, reservoirs and coastal areas in the wild. Such devices are usually powered by batteries combined with solar power, and their energy consumption management level directly determines the continuous working life of the device and the reliability of the entire monitoring network. In the context of frequent extreme weather events, real-time and reliability of water level monitoring are required, which forms a prominent contradiction with the limited energy supply of the device.

[0003] At present, the power consumption control of water level monitoring devices mainly relies on the following traditional technologies: first, periodic sampling and sleep mechanism with fixed time interval, that is, the device is awakened at a predetermined time point for data collection and transmission, and the rest of the time is in low-power sleep state. This method is simple to implement, but lacks flexibility and cannot adaptively adjust according to the actual hydrological conditions. In the long dry season, this fixed cycle will cause waste of energy; while in emergency situations such as heavy rain and floods, key data may be missed due to insufficient sampling frequency, resulting in ineffective warning. Second, trigger control based on a single threshold, for example, when the water level exceeds a certain preset limit, the device increases the working frequency. This method responds to environmental changes with a lag, and the threshold setting usually relies on historical experience, making it difficult to cope with complex and variable climate patterns and basin characteristics.

[0004] In recent years, with the development of Internet of Things technology, some methods have appeared that attempt to optimize power consumption through environmental parameters. For example, some solutions propose to fine-tune the device's working mode according to environmental temperature or light intensity. However, these methods mostly only consider a few isolated parameters and fail to systematically integrate macro environmental background, real-time running state of the device itself, and long-term hydrological regularity. Lack of a multi-dimensional context awareness and decision-making framework, resulting in power consumption strategies often being locally optimized, unable to achieve optimal allocation of global energy consumption while ensuring the reliability of monitoring tasks. SUMMARY

[0005] To solve the problems of poor flexibility, insufficient environmental adaptability and lack of forward-looking verification in the prior art, the purpose of the present application is to provide a water level monitoring device power consumption autonomous regulation method and system using context-aware, by constructing a multi-level context-aware mechanism, dynamically evaluating device operation value and environmental risk, intelligently constructing device internal power consumption topology structure, and combining digital twin technology to verify the regulation scheme, realizing adaptive balance of power consumption and monitoring performance.

[0006] The technical solution of the present application is as follows:

[0007] One of the purposes of the present application is to provide a water level monitoring device power consumption autonomous regulation method using context-aware, comprising:

[0008] S100: Construct a multi-level context-aware layer, access external information sources through a low-power mode, extract context-aware labels representing macro-environmental background, and form a context-aware model in combination with local historical hydrological rules;

[0009] S200: Real-time collection of multi-dimensional operating state data, according to the coupling relationship between multi-dimensional operating state data and the context-aware model, execute situation-enhanced power consumption value dynamic evaluation, judge the power consumption regulation level that matches the monitoring stage and environmental risk level of the device;

[0010] S300: According to the power consumption regulation level and the current context-aware model, the sensor, processor, communication unit and storage unit are regarded as energy consumption nodes that can be flexibly combined, the power consumption topology structure of the internal functional modules of the water level monitoring device is constructed, and according to the demand difference of the power consumption topology structure under different situations, a scheme group for reducing power consumption is generated;

[0011] S400: Based on the context-aware model and real-time operating data, a device digital twin is constructed for the water level monitoring device, the scheme group for reducing power consumption is loaded, the system resilience index of the scheme group under the risk situation is quantitatively evaluated, and the optimal power consumption reduction scheme is determined;

[0012] S500: The optimal power consumption reduction scheme is deployed to each node in the actual water level monitoring network for execution, and the environmental changes and strategy execution effect are continuously monitored, when the environmental mode changes or the strategy efficiency decreases, automatically trigger and return to step S100.

[0013] As a further option of the present application, the context-aware model is: ; wherein, is a situation modeling function, is a context-aware model, is a context-aware label, and H is local historical hydrological data.

[0014] As a further choice of the present application, in step S100, the context situation modeling forming step comprises:

[0015] Building context situation label based on multi-source external information , and fusing local historical hydrological data H; wherein the context situation label is , wherein M is the dimension of the situation label, represents the value of the jth situation label at time t;

[0016] Adopting a weighted moving average method to perform time series smoothing processing on the situation label to extract a trend value ;

[0017] Infering the current environment state in combination with a hidden Markov model , wherein the state space is ; the environment state Calculating the probability of being in each state at the current time through a forward-backward algorithm, and taking the maximum probability state as ;

[0018] Outputting an enhanced state vector containing a situation trend, an environment state and a state confidence through the context situation model, wherein the context situation model output is an enhanced state vector, and the enhanced state vector is: ; wherein is the state confidence.

[0019] As a further choice of the present application, in step S200, the situation-enhanced power consumption value dynamic evaluation comprises:

[0020] Performing a situation-enhanced coupling relationship analysis based on the enhanced state vector output by the context situation model and the multi-dimensional running state data of the device, to form a coupling relationship matrix R;

[0021] Calculating a power consumption value index through a power consumption value function, wherein the power consumption value function is defined as: ; wherein, is the power consumption value index, is the current available energy, is the total energy capacity, is the environmental risk level, is the task urgency, and α, β, γ are weight coefficients satisfying ;

[0022] According to the comparison between the power consumption value index and a preset threshold value, determining the power consumption regulation level.

[0023] As a further choice of the present application, the situation-enhanced coupling relationship analysis comprises:

[0024] The situation label The environmental state The unified code is a numerical feature vector , wherein is an indicator function, is a numerical feature vector, is a state confidence; the numerical feature vector has a dimension of M+4;

[0025] For each state indicator The context feature , the correlation coefficient is obtained by providing calculation through the Pearson correlation coefficient ;

[0026] Form The coupling relationship matrix has a dimension of M+4; .

[0027] As a further selection of the present application, the power consumption regulation level : ; wherein, and are threshold parameters.

[0028] As a further selection of the present application, in step S300, the power consumption topology structure of the internal function module of the water level monitoring device is constructed, including:

[0029] The internal function module of the device is abstracted as an energy consumption node graph, wherein the node set includes sensor nodes, processor nodes, communication nodes and storage nodes, and the edge set represents the energy consumption dependency relationship between nodes;

[0030] A power consumption topology objective function is defined to minimize the total power consumption while meeting performance constraints, and the objective function is: ; the constraint conditions include: ; wherein, is the minimum total power consumption, is the minimum performance requirement, and G is the energy consumption node graph, i.e. , wherein V is the node set, E is the edge set, , , respectively represent the power consumption, performance contribution and state variable associated with each node in the node set V.

[0031] As a further selection of the present application, the power consumption reduction scheme group is generated by using a greedy algorithm and a constraint solver, specifically including:

[0032] Based on the current power consumption topology, a number of feasible solutions are randomly generated;

[0033] Using a greedy search, starting from the initial solution, gradually adjusting the node state to reduce power consumption;

[0034] verify whether each solution satisfies the constraint condition;

[0035] generate a group of solutions for reducing power consumption each solution corresponds to a topology configuration.

[0036] As a further option of the present application, in step S400, the system resilience index is quantified by the following formula: ; wherein, is the actual power consumption at time t, is the nominal power consumption, is the decay coefficient, and is the simulation start and end time;

[0037] By comparing the resilience indexes of each solution, the optimal solution is selected in descending order.

[0038] The second object of the present application is to provide a water level monitoring device power consumption autonomous regulation system using context-awareness, comprising:

[0039] The context-awareness and modeling module is used to build a multi-level context-awareness layer, access external information sources through a low-power mode, extract context situation labels representing macro-environmental backgrounds, and form a context situation model in combination with local historical hydrological rules.

[0040] The dynamic evaluation and decision module is used to collect multi-dimensional running state data in real time, perform situation-enhanced power consumption value dynamic evaluation according to the coupling relationship between the multi-dimensional running state data and the context situation model, and judge the power consumption regulation level that matches the monitoring stage and the environmental risk level of the device.

[0041] The power consumption solution generation module is used to regard the sensor, processor, communication unit and storage unit as energy consumption nodes that can be flexibly combined, construct the power consumption topology structure of the internal function modules of the water level monitoring device according to the power consumption regulation level and the current context situation model, generate a group of solutions for reducing power consumption according to the demand differences of the power consumption topology structure under different situations.

[0042] The digital twin and optimization module is used to construct a device digital twin for the water level monitoring device based on the context situation model and real-time running data, load the group of solutions for reducing power consumption, determine the optimal power consumption reduction solution by quantitatively evaluating the system resilience index of the group of solutions under the risk situation.

[0043] A strategy execution and feedback module is configured to deploy the optimal power consumption reduction scheme to each node in the actual water level monitoring network for execution, and continuously monitor environmental changes and the effect of strategy execution, and when detecting a change in environmental patterns or a decline in strategy effectiveness, automatically trigger the context-aware and modeling module to start working again.

[0044] The application has the following advantages:

[0045] The application systematically integrates macro environmental background and local historical hydrological rules by constructing a multi-level context-aware layer, breaking through the limitations of traditional methods that rely on single or fixed parameters. This method can dynamically capture the temporal changes and spatial characteristics of environmental patterns, realizing the transition from "passive response" to "forward-looking perception", thereby providing a precise and comprehensive context awareness basis for power consumption regulation.

[0046] Through the context-enhanced power consumption value dynamic evaluation mechanism and dynamic power consumption topology construction theory, the multi-dimensional running state inside the device and the external environmental background are deeply integrated for decision-making. This not only realizes the precise matching of power consumption regulation level, monitoring task, and environmental risk, but also realizes the fine management of energy consumption through modular topology optimization, solving the balance problem between global energy efficiency optimization and task reliability guarantee of traditional methods.

[0047] By constructing a device digital twin and quantifying system resilience indicators, the scheme completes the verification and optimization of the regulation strategy in the virtual space, ensuring the high robustness of the strategy in real complex environments. Combined with the continuous monitoring and closed-loop optimization mechanism after deployment, the system has the full life cycle adaptive ability of self-awareness, self-decision, self-verification, and self-optimization, significantly improving the continuous working life and overall reliability of the water level monitoring network under extreme conditions. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 The overall flowchart of the water level monitoring device power consumption autonomous regulation method using context-awareness;

[0049] Figure 2 The detailed flowchart of step S100 of the water level monitoring device power consumption autonomous regulation method using context-awareness;

[0050] Figure 3 The detailed flowchart of step S200 of the water level monitoring device power consumption autonomous regulation method using context-awareness;

[0051] Figure 4 The detailed flowchart of step S300 of the water level monitoring device power consumption autonomous regulation method using context-awareness;

[0052] Figure 5A detailed flowchart of step S400 of the method for autonomous power consumption regulation of a water level monitoring device using context-awareness;

[0053] Figure 6 A detailed flowchart of step S500 of the method for autonomous power consumption regulation of a water level monitoring device using context-awareness. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0055] As a key component of water conservancy informatization system, the water level monitoring device is directly related to the long-term stable operation of the device in the field environment. In the complex and changeable hydrological environment, power consumption management faces challenges such as dynamic changes in environmental context, weak correlation of multi-dimensional state data, poor adaptability of power consumption regulation strategy, etc. The theoretical basis of the present application is built on four pillars of context situation modeling theory, power consumption value dynamic evaluation theory, topology construction theory and system resilience quantification theory. The precise extraction of environmental background is realized through the construction of a multi-level context situation perception layer, the regulation level is determined by using the situation-enhanced power consumption value dynamic evaluation, the scheme group for reducing power consumption is generated by combining the dynamic power consumption topology construction, and the system resilience index is quantitatively evaluated by the digital twin to ensure the regulation reliability.

[0056] The core theories are as follows:

[0057] Firstly, the context situation perception layer can be modeled as a multi-source information fusion system, and its state is represented as: ; wherein, is the context situation label, M is the dimension of the situation label, represents the value of the jth situation label at time t, such as environmental temperature, rainfall intensity, water level change rate, etc. The context situation model is constructed by historical hydrological law and real-time external information source, and is defined as: ; wherein, H is the local historical hydrological data, is a situation modeling function, which combines weighted moving average and hidden Markov model to capture the time sequence dependence of the environment mode.

[0058] Secondly, to quantify the power consumption regulation value, the present application introduces a situation-enhanced power consumption value dynamic evaluation mechanism. Let the multi-dimensional running state data of the device be , wherein N is the number of state indicators, such as battery voltage, CPU utilization, communication load, etc. The power consumption value function is defined as: ; wherein, is the power consumption value index, is the current available energy, is the total energy capacity, is the environmental risk level, is the task urgency. a, b, g are weight coefficients, satisfying . By comparing with the preset threshold, the power consumption regulation level is determined:

[0059] ;

[0060] wherein, and are threshold parameters, level 1 represents a high power consumption mode, level 2 represents a medium power consumption mode, and level 3 represents a low power consumption mode.

[0061] Thirdly, to realize the construction of the power consumption topology structure, the internal functional modules of the device are abstracted into an energy consumption node graph , wherein V is a node set, such as a sensor, a processor, a communication unit, and a storage unit, and E is an edge set, representing the energy consumption dependency relationship between the modules. The goal of the power consumption topology construction is to minimize the total power consumption while satisfying the performance constraints: ; wherein G is the constructed topology, is the power consumption of the node , and is a binary variable, with 1 representing activation and 0 representing dormancy. The constraint conditions include: ; wherein, is the performance contribution of the node , and is the minimum performance requirement. By using a greedy algorithm and constraint solving, a set of power consumption reduction schemes is generated, wherein each scheme corresponds to a topology configuration.

[0062] Finally, to evaluate the reliability of the scheme set, a device digital twin T is constructed to simulate the behavior of the actual device under risk scenarios. The system resilience index is defined as the ability of the device to maintain functionality under failure or stress, and the calculation formula is: ; wherein, is the actual power consumption, is the nominal power consumption, is the decay coefficient, reflecting the influence of time on resilience. By comparing the of each scheme, the optimal scheme is selected: . This mechanism ensures that the regulation strategy has high robustness in complex environments.

[0063] The above theoretical framework provides a solid mathematical foundation for the application, ensuring the precision, adaptability and reliability of the autonomous power consumption regulation. The specific embodiments of the application will be described in detail below.

[0064] Embodiment one;

[0065] Please refer to Figure 1 , which shows a kind of water level monitoring equipment power consumption autonomous regulation method provided by the context-awareness of the embodiment of the application, the method comprises:

[0066] S100: build multi-level context-awareness layer and context model;

[0067] S200: collect multi-dimensional operation data, evaluate power consumption value regulation level;

[0068] S300: construct power consumption topology according to regulation level, generate power reduction scheme group;

[0069] S400: construct digital twin, evaluate resilience and select optimal power reduction scheme;

[0070] S500: deploy optimal scheme, monitor and adaptively trigger regulation iteration.

[0071] The specific scheme is as follows:

[0072] In a kind of water level monitoring equipment power consumption autonomous regulation method using context-awareness, S100 realizes the multi-source information fusion of macro-environment background and context modeling. Multi-level context-awareness layer is accessed to external information source by low-power communication protocol, and combines local historical hydrological law, forms dynamic updated context model, provides environmental background support for subsequent power consumption value evaluation and topology construction.

[0073] Please refer to Figure 2 , which shows a kind of water level monitoring equipment power consumption autonomous regulation method S100 using context-awareness of the embodiment of the application, its content includes:

[0074] S110: low-power access to multi-source external information, collect macro-environment data.

[0075] Context-awareness layer is accessed to external information source by low-power wide area network, and macro-environment data is acquired in real time. External information source includes meteorological data, hydrological bulletin, topographic map, satellite remote sensing image, etc., covers temperature, rainfall, evaporation, water level, flow and other multi-dimensional indexes.

[0076] Specifically, external information access includes: meteorological data acquisition, hydrological data acquisition, geographic data acquisition, remote sensing data acquisition.

[0077] In an alternative embodiment, the external information access employs edge gateway proxy technology, data preprocessing and compression are performed at the local gateway, and only key feature values are uploaded to reduce communication power consumption.

[0078] S120: Extract context labels to form standardized context labels.

[0079] Extract context labels representing the macro-environmental background from the original external information to form standardized context labels where M is the dimension of the context label, denotes the value of the jth context label at time t.

[0080] In a possible embodiment, the extraction of context labels includes the following steps:

[0081] 1) Remove outliers and missing values, and use interpolation to complete them;

[0082] 2) Map the original data to the interval [0, 1] to eliminate the influence of dimension;

[0083] 3) According to the field knowledge, the continuous values are discretized into context levels, for example, the rainfall intensity is divided into "no rain", "light rain", "moderate rain", and "heavy rain";

[0084] 4) Encode the discrete labels into numerical vectors for model processing.

[0085] S130: Analyze local historical hydrological data to form a local hydrological law library.

[0086] Combine the historical hydrological data of the device deployment point to analyze the water level change law, seasonal characteristics and extreme event frequency, and form a local hydrological law library, i.e. local historical hydrological data H. The historical hydrological data includes water level time series, flow records, flood events, drought cycles, etc.

[0087] S140: Construct a context model using weighted moving average and hidden Markov model.

[0088] Based on the context labels and the local historical hydrological data H, a context model is constructed. The model uses a combination of weighted moving average and hidden Markov model to capture the time dependence of the environment pattern.

[0089] In a possible embodiment, the context model is: ; where, is the context modeling function.

[0090] Based on the above embodiment, the context situation model calculation method is:

[0091] For each context label Temporal smoothing processing is performed, and a weighted moving average method is used to calculate the trend value ;

[0092] In combination with local historical hydrological data H, the current environment state is inferred through a hidden Markov model . The state space is defined as , and the transition probability matrix is obtained by statistical analysis of historical hydrological events. The probability of being in each state at the current time is calculated by the forward-backward algorithm, and the state with the maximum probability is taken as .

[0093] The context situation model output is an enhanced state vector: ;

[0094] Wherein is the state confidence.

[0095] In a power consumption autonomous regulation method for a water level monitoring device using context situation awareness, S200 performs a situation-enhanced power consumption value dynamic evaluation based on real-time collected multi-dimensional operating state data and a context situation model, judges the power consumption regulation level that matches the monitoring stage and the environmental risk level of the device. This step deeply integrates the device operating state and the environmental background, and realizes accurate decision-making of power consumption regulation.

[0096] Please refer to Figure 3 , which shows a flowchart of an example of a power consumption autonomous regulation method S200 for a water level monitoring device using context situation awareness, which includes:

[0097] S210: Collecting the operating state data of each module in the device.

[0098] Real-time collection of operating state data of each module in the water level monitoring device forms a state vector , wherein N is the number of state indicators.

[0099] Specifically, the multi-dimensional operating state data includes: energy state data, calculation state data, communication state data, and storage state data.

[0100] In a possible implementation, the multi-dimensional operating state data collection uses a lightweight agent program embedded in the device firmware, which runs at a low sampling frequency to reduce its own power consumption.

[0101] S220: Analyzing the coupling relationship between the situation and the operating data to form a coupling relationship matrix.

[0102] The enhanced state vector output by the context situation model constructed based on S140 , and the multi-dimensional running state data of the device , perform situation-enhanced coupling relationship analysis to form a coupling relationship matrix R.

[0103] In one possible implementation, the coupling relationship analysis includes:

[0104] The situation label and the discrete environmental state are uniformly encoded into a numerical feature vector , where is an indicator function. The vector dimension is M+4.

[0105] For each state indicator and situation feature , the correlation coefficient is obtained by providing calculation through the Pearson correlation coefficient.

[0106] Finally, a dimension coupling relationship matrix R is formed: .

[0107] The coupling relationship matrix R quantifies the dynamic correlation strength between the internal running state of the device and the external environmental situation.

[0108] S230: Calculate the power consumption value index in combination with the coupling relationship matrix.

[0109] The coupling relationship matrix R calculates the power consumption value index to reflect the benefit-risk ratio of current power consumption. The power consumption value function is defined as: ; where is the power consumption value index, is the current available energy, is the total energy capacity, is the environmental risk level, is the task urgency. α, β, γ are weight coefficients, satisfying .

[0110] In one possible implementation, the power consumption value dynamic evaluation step through the power consumption value function includes:

[0111] 1) Obtain , , , ; where and are read from the battery management unit; is the environmental state in the context situation model , mapped to risk value (01); Obtained based on the priority and timing requirements of the monitoring tasks;

[0112] 2) Dynamically adjust α, β, and γ based on the coupling matrix R; for example, increase β when the environmental risk is high.

[0113] 3) Real-time calculation using power consumption value function .

[0114] S240: Compare power consumption value index to determine power consumption control level.

[0115] According to the power consumption value index The power consumption control level is determined by comparing it with a preset threshold. : ;in, and The threshold parameter represents the high power consumption mode, the medium power consumption mode, and the low power consumption mode. Level 1 represents the high power consumption mode, Level 2 represents the medium power consumption mode, and Level 3 represents the low power consumption mode.

[0116] In a context-aware method for autonomous power consumption control in water level monitoring equipment, the S300, based on the power consumption control level and the current context model, treats sensors, processors, communication units, and storage units as flexibly combinable energy-consuming nodes, constructs a power consumption topology for the internal functional modules of the equipment, and generates a set of power-saving solutions. This step achieves refined power consumption management through topology optimization.

[0117] Please refer to Figure 4 The diagram illustrates a flowchart of an exemplary method for autonomous power consumption control of a water level monitoring device using a context-aware approach, S300, which includes:

[0118] S310: Abstract the device functional modules into an energy consumption node diagram.

[0119] Abstracting the internal functional modules of the equipment into an energy consumption node diagram Where V is the node set, including sensor nodes, processor nodes, communication nodes, and storage nodes; E is the edge set, representing the energy consumption dependencies between nodes, such as data flow and control flow.

[0120] S320: Defines the power consumption topology objective function and performance constraints.

[0121] Energy consumption node graph of power consumption topology Internal node attribute definition, that is, defining each node in the node set V. Related power consumption Performance contribution State variables ,in, Represents activation. represent hibernation.

[0122] The goal of the power consumption topology is to minimize the total power consumption while satisfying the performance constraints. The objective function is defined based on the energy consumption node graph The objective function is defined based on the energy consumption node graph The constraints include: ; where, is the minimum performance requirement, determined by the monitoring task.

[0123] S330: Generate a set of power reduction schemes that satisfy the constraints using a greedy algorithm.

[0124] A greedy algorithm and a constraint solver are used to generate topology configuration schemes that satisfy the constraints. The construction algorithm takes into account the power consumption regulation level and the context situation model to adapt to different environments.

[0125] In one possible implementation, the construction algorithm step includes:

[0126] 1) Based on the current power consumption topology, a number of feasible solutions are randomly generated;

[0127] 2) A greedy search is used to start from the initial solution and gradually adjust the node state to reduce power consumption;

[0128] 3) Verify whether each solution satisfies the constraint conditions;

[0129] 4) Generate a set of power reduction schemes Each scheme corresponds to a topology configuration.

[0130] In a context situation-aware power consumption autonomous regulation method for water level monitoring equipment, S400 constructs a device digital twin for the water level monitoring equipment based on the context situation model and real-time operation data, loads a set of power reduction schemes, determines the optimal power reduction scheme by quantitatively evaluating the system resilience indicators of the set of schemes in the risk situation. This step realizes controllable verification of the scheme through digital twin technology.

[0131] Please refer to Figure 5 , which shows a flowchart of one example of a context situation-aware power consumption autonomous regulation method S400 for water level monitoring equipment, the content of which includes:

[0132] S410: Construct a digital twin based on device parameters and situation model.

[0133] Based on the device physical model, operation logic and context model, a high-fidelity digital twin T is constructed. The digital twin simulates the dynamic behavior of the actual device, including power consumption, performance, failure modes, etc.

[0134] In one possible implementation, the digital twin construction includes:

[0135] 1) Extract parameters from the actual device, such as battery capacity, CPU frequency, communication protocol;

[0136] 2) Inject the context model into the twin, simulating the real environment;

[0137] 3) Simulate device operation through discrete event simulation or system dynamics model, forming a digital twin.

[0138] S420: Load scenario group simulation, collect data under risk scenarios.

[0139] After screening, the scenario group is loaded into the digital twin to simulate the execution effect of each scenario under the risk scenario. The risk scenario includes extreme weather, hardware failure, network interruption, etc. The power consumption, performance, and failure events during simulation are recorded, and the simulation data is collected for resilience index calculation.

[0140] S430: Calculate system resilience index based on simulation data.

[0141] Based on the simulation data, the system resilience index of each scenario is calculated, reflecting the maintenance ability of the scheme under pressure. The resilience index is defined as: ; wherein, is the actual power consumption at time t, is the nominal power consumption, is the decay coefficient, and is the simulation start and end time.

[0142] S440: Compare the resilience index to determine the optimal power reduction scheme.

[0143] Compare the resilience index of each scheme , arrange the schemes in descending order of , and select the optimal scheme , that is: .

[0144] ​In a context-aware water level monitoring device power consumption autonomous regulation method, S500 deploys the optimal power reduction scheme to each node in the actual water level monitoring network for execution, and continuously monitors environmental changes and policy execution effects. When detecting changes in environmental patterns or declining policy effectiveness, automatically trigger and return to step S100. This step realizes regulation and long-term adaptation.

[0145] Please refer to Figure 6 , which shows a flowchart of an example of a context-aware water level monitoring device power consumption autonomous regulation method S500, which includes:

[0146] S510: Convert optimal scheme into instruction deployment and monitor execution in multiple dimensions.

[0147] Convert the optimal power reduction scheme verified by the digital twin into executable instructions and establish a multi-dimensional monitoring system.

[0148] First, convert the optimal power reduction scheme into a set of device-recognizable control instructions, including sensor sampling frequency adjustment, processor operating mode switching, communication protocol stack reconstruction, and storage strategy update. The deployment process uses a hierarchical progressive approach, with priority given to instruction verification at the edge gateway node, followed by batch delivery to terminal devices through LoRaWAN / NBIoT networks. To ensure deployment reliability, digital signatures and rollback mechanisms are introduced. When instruction verification fails, automatically restore to the previous stable version.

[0149] At the device level during the policy execution phase, real-time power consumption data for each functional module is collected to obtain the actual power saving rate index. At the network level during the policy execution phase, link quality index and data packet delivery rate are collected through the SDN controller to evaluate the optimal power reduction scheme execution effect. At the platform level during the policy execution phase, a streaming processing pipeline is built based on ApacheFlink to perform real-time aggregation analysis on 12 core indicators such as device health, task completion rate, and environmental adaptability. In particular, for extreme scenarios such as heavy rain and floods, an LSTM-based anomaly detection module is set up. When the water level change rate exceeds the threshold , automatically trigger the emergency sampling mode to maintain controllable power consumption while ensuring data integrity.

[0150] S520: Multi-condition triggered backtracking to achieve policy optimization.

[0151] By building a multi-condition triggering mechanism and a knowledge evolution system, autonomous iterative optimization of the regulation strategy is achieved.

[0152] ​In one possible implementation, by establishing a multi-dimensional perception-based trigger judgment process, the rollback process is immediately started when any of the following conditions is met:

[0153] 1) The environment mode has an essential change, which is represented by the transition probability of the environment state in the context situation model is lower than the historical average by 2 standard deviations;

[0154] 2) The strategy effectiveness continues to decline, i.e., the power saving rate is lower than 25% for 3 consecutive monitoring periods.

[0155] When the trigger condition is met, the system automatically falls back to S100 to reinitialize.

[0156] The present application has been comprehensively verified in the Yangtze River Basin water level monitoring network, which contains 200 monitoring nodes, and the monitoring indicators include water level, flow, rainfall, temperature, etc. During the implementation process, the following specific configurations are used:

[0157] Context perception layer: access to public data of China Meteorological Administration and Ministry of Water Resources, use LoRaW transmission, sampling frequency 1 / minute;

[0158] Data processing layer: configure edge server, perform context modeling and value evaluation, use Python and TensorFlowLite;

[0159] Digital twin layer: build twin based on AWSIoTGreengrass, simulation software is AnyLogic;

[0160] Strategy execution layer: integrate Huawei OceanConnect IoT platform, support remote deployment and monitoring.

[0161] In the test phase, 10 types of risk situations such as heavy rain, drought, and device failure are simulated, each type of situation is repeated 15 times. The test results show that:

[0162] The average power saving rate is 35.7%, and the highest is 50% in low-risk situations;

[0163] The system resilience index is improved by 42.3%, and the device continues to run in extreme environments for 60% longer;

[0164] The average response time of regulation and control decision is 8 seconds, which meets the real-time requirement;

[0165] The adaptive mechanism reduces the long-term operation and maintenance cost by 28%.

[0166] ​In particular, in a real rainstorm event, the application successfully predicts the water level rising trend, switches to low-power mode in advance, prolongs the battery life by 72 hours, and avoids data loss. Digital twin evaluation shows that the system resilience index reaches 0.89, with a confidence level of 96.8%, verifying the regulation reliability.

[0167] Embodiment two;

[0168] A water level monitoring device power consumption autonomous regulation system using context-awareness, comprising:

[0169] A context-awareness and modeling module for building a multi-level context-awareness layer, accessing external information sources through a low-power mode, extracting context tags representing macro-environmental background, and forming a context model in combination with local historical hydrological rules;

[0170] A dynamic evaluation and decision-making module for real-time collection of multi-dimensional operating state data, performing context-enhanced power consumption value dynamic evaluation based on the coupling relationship between multi-dimensional operating state data and context model, and determining the power consumption regulation level matching the monitoring stage and environmental risk level of the device;

[0171] A power consumption scheme generation module for constructing a power consumption topology structure of the internal functional modules of the water level monitoring device by regarding sensors, processors, communication units and storage units as flexible combination energy consumption nodes based on the power consumption regulation level and the current context model, and generating a scheme group for reducing power consumption according to the demand differences of the power consumption topology structure under different contexts;

[0172] A digital twin and optimization module for constructing a device digital twin for the water level monitoring device based on the context model and real-time operating data, loading the scheme group for reducing power consumption, and determining the optimal power consumption reduction scheme by quantitatively evaluating the system resilience index of the scheme group under risk scenarios;

[0173] A strategy execution and feedback module for deploying the optimal power consumption reduction scheme to each node in the actual water level monitoring network for execution, and continuously monitoring environmental changes and strategy execution effects, and automatically triggering the context-awareness and modeling module to start working again when detecting changes in environmental patterns or a decline in strategy effectiveness.

[0174] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0175] Although preferred embodiments of this application 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 the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0176] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for autonomous power consumption control of water level monitoring equipment with context awareness, characterized in that: include: S100: Constructs a multi-level context awareness layer, accesses external information sources in a low-power manner, extracts context labels that represent the macro-environmental background, and combines them with local historical hydrological patterns to form a context model. S200: Real-time acquisition of multi-dimensional operating status data; based on the coupling relationship between multi-dimensional operating status data and contextual model, perform context-enhanced dynamic power consumption value assessment to determine the power consumption control level that matches the monitoring stage of the device and the environmental risk level. S300: Based on the power consumption control level and the current context scenario model, the sensor, processor, communication unit and storage unit are regarded as energy consumption nodes that can be flexibly combined. The power consumption topology of the internal functional modules of the water level monitoring equipment is constructed. According to the differences in the power consumption topology requirements under different scenarios, a group of schemes to reduce power consumption is generated. S400: Based on the contextual model and real-time operating data, a digital twin of the water level monitoring equipment is built, a group of power reduction solutions are loaded, and the optimal power reduction solution is determined by quantitatively evaluating the system resilience index of the solution group under risk scenarios. S500: Deploy the optimal power reduction scheme to each node in the actual water level monitoring network and continuously monitor environmental changes and policy execution effects. When a change in environmental mode or a decrease in policy performance is detected, automatically trigger and return to execution step S100. In step S100, the contextual model is: ;in, Modeling functions for the context For contextual models, For contextual labels, For local historical hydrological data; In step S100, the contextual modeling formation step includes: Contextual labels are constructed based on multi-source external information. And integrate local historical hydrological data Among them, the context label is ,in For contextual labeling dimension, Indicates the first A contextual label in time The possible values ​​of ; A weighted moving average method is used to smooth the context labels over time in order to extract trend values. ; Inferring the current environmental state using hidden Markov models The state space is: Environmental conditions The probability of being in each state at the current moment is calculated using a forward-backward algorithm, and the state with the highest probability is selected as the [state name]. ; The contextual context model outputs an enhanced state vector containing contextual trends, environmental states, and state confidence. The enhanced state vector output by the contextual context model is as follows: ;in, State confidence; In step S200, the context-enhanced power consumption value dynamic assessment includes: Based on the enhanced state vector output by the contextual model and the multi-dimensional operating state data of the equipment, a context-enhanced coupling relationship analysis is performed to form a coupling relationship matrix. ; The power consumption value index is calculated using the power consumption value function, where the power consumption value function is defined as follows: ;in, The power consumption value index For currently available energy, Total energy capacity Environmental risk level, To assess the urgency of the task, , , Let be the weighting coefficient, satisfying ; The power consumption control level is determined by comparing the power consumption value index with a preset threshold. In step S400, each scheme The system resilience index is quantified by the following formula: ;in, For time Actual power consumption This is the nominal power consumption. The attenuation coefficient is... and The simulation start and end times are defined; by comparing the resilience indices of each scheme, the optimal scheme is selected in descending order.

2. The method for autonomous power consumption control of water level monitoring equipment using context-awareness as described in claim 1, characterized in that, The context-enhanced coupling analysis includes: context labels With environmental conditions Unified encoding as numerical feature vectors ,in For indicator functions, It is a numerical feature vector. State confidence; numerical feature vector Dimensions ; For each state indicator Contextual features The correlation coefficient is obtained by using the Pearson correlation coefficient. ; form 3D coupling matrix: .

3. The method for autonomous power consumption control of water level monitoring equipment using context-awareness as described in claim 1, characterized in that, The power consumption control level : ;in, and This is the threshold parameter.

4. The method for autonomous power consumption control of water level monitoring equipment using context-awareness as described in claim 1, characterized in that, In step S300, the power consumption topology of the internal functional modules of the constructed water level monitoring device includes: The internal functional modules of the device are abstracted into an energy consumption node graph, where the node set includes sensor nodes, processor nodes, communication nodes and storage nodes, and the edge set represents the energy consumption dependency relationship between nodes; Define a power consumption topology objective function to minimize total power consumption while satisfying performance constraints. The objective function is: The constraints include: ;in, To minimize total power consumption, For minimum performance requirements, This is an energy consumption node diagram, i.e. ,in For a set of nodes, For edge set, , , Representing node sets respectively Each node in Related power consumption, performance contribution, and state variables.

5. The method for autonomous power consumption control of water level monitoring equipment using context-awareness as described in claim 4, characterized in that, In step S300, the generation of the power-reducing scheme group is performed using a greedy algorithm and a constraint solver, specifically including: Based on the current power consumption topology, several feasible solutions are randomly generated; A greedy search algorithm is used to gradually adjust the node states starting from the initial solution in order to reduce power consumption. Verify that each solution satisfies the constraints; Generate a set of solutions to reduce power consumption Each plan It corresponds to one topology configuration.

6. A control system employing a context-aware autonomous power consumption control method for water level monitoring equipment according to any one of claims 1-5, characterized in that, include: The context awareness and modeling module is used to build a multi-level context awareness layer. It accesses external information sources in a low-power manner, extracts contextual labels that represent the macro-environmental background, and combines them with local historical hydrological patterns to form a contextual model. The dynamic assessment and decision-making module is used to collect multi-dimensional operating status data in real time. Based on the coupling relationship between the multi-dimensional operating status data and the contextual model, it performs a context-enhanced dynamic assessment of power consumption value and determines the power consumption control level that matches the monitoring stage of the device and the environmental risk level. The power consumption scheme generation module is used to construct the power consumption topology of the internal functional modules of the water level monitoring equipment, based on the power consumption control level and the current context scenario model, treating the sensor, processor, communication unit and storage unit as energy consumption nodes that can be flexibly combined. According to the different power consumption topology requirements under different scenarios, it generates a group of schemes to reduce power consumption. The digital twin and optimization module is used to build a digital twin of the water level monitoring equipment based on the contextual model and real-time operating data, load a group of power reduction solutions, and determine the optimal power reduction solution by quantitatively evaluating the system resilience index of the solution group under risk scenarios. The strategy execution and feedback module is used to deploy the optimal power reduction scheme to each node in the actual water level monitoring network and continuously monitor environmental changes and strategy execution effects. When a change in environmental mode or a decrease in strategy performance is detected, the context awareness and modeling module is automatically triggered to restart.

Images