Method and system for recommending optimal maintenance scheme of Internet of Things

Abstract

The invention provides a recommendation method for an optimal maintenance scheme of the Internet of Things. The recommendation method comprises the following steps: step 1, acquiring system input; step 2, performing WSN simulation and cost evaluation based on the system input; and step 3, performing optimal maintenance scheme search under budget constraint by adopting an intelligent search algorithm. According to the method, the automatic recommendation of the maintenance module and the parameter combination configuration thereof can be realized by establishing an optimization model between the overall maintenance cost and the system reliability under the budget constraint aiming at factors such as network scales, node distribution, fault types and occurrence probabilities, energy consumption models and the like in different application scenes. According to the method, typical maintenance function modules such as remote restart, heartbeat detection and firmware updating serve as candidate components capable of being started as required, key parameters such as an energy early warning threshold value and heartbeat frequency are combined, the actual effect of a maintenance strategy is evaluated based on system modeling and simulation analysis, and therefore the optimal maintenance scheme is screened out.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) maintenance technology, and in particular to a method and system for recommending optimal IoT maintenance solutions. Background Technology

[0002] In recent years, with the development of IoT technology, sensing devices have been widely deployed in industrial monitoring, smart cities, environmental sensing, and other scenarios, forming complex wireless sensor network (WSN) systems. These systems typically feature a large number of nodes, complex deployment environments, and limited resources (such as power, computing power, and communication bandwidth), making them prone to communication anomalies, sudden increases in energy consumption, and node failures during operation. These problems not only increase system maintenance costs but also pose greater challenges to traditional maintenance strategies.

[0003] In the field of computer system and network maintenance, several general maintenance strategies have been proposed, including corrective maintenance for post-fault repair, preventive maintenance performed periodically, and predictive maintenance relying on status monitoring. Predictive maintenance, in particular, relies on sensing infrastructure such as WSNs to monitor and analyze system status in real time, in order to detect potential faults early. However, it is worth noting that WSNs themselves, as critical support systems in the operating environment, may also face problems such as low reliability, resource constraints, and maintenance complexity, making traditional maintenance strategies difficult to apply directly.

[0004] While existing computer system and network maintenance strategies are relatively mature in traditional environments, their direct application to WSN systems still faces numerous challenges. First, corrective maintenance relies on reactive response mechanisms after a failure occurs. In large-scale, distributed WSN systems, delayed responses often lead to fault propagation and system-level anomalies, even resulting in irrecoverable critical data. Second, while predictive maintenance boasts a high level of intelligence, its reliance on state acquisition, modeling, and analysis requires significant computing and communication resources, making it difficult to deploy on energy-constrained edge nodes and resulting in high implementation costs. In contrast, preventative maintenance, a widely used strategy in current WSNs, typically achieves periodic maintenance through the configuration of modules such as remote restart and heartbeat detection. However, its module selection and parameter configuration largely depend on manual experience, lacking systematic modeling and evaluation of system operating status, fault characteristics, and resource constraints. This static configuration approach struggles to adapt to the maintenance needs of complex application scenarios, easily leading to resource waste and maintenance failures.

[0005] Furthermore, although existing research has made some progress in the reliability and fault tolerance mechanisms of IoT systems, attention to the maintainability of WSN systems remains limited. Related work largely focuses on static indicator modeling, and there is a lack of a method and technical architecture that can, at the design stage, comprehensively evaluate application scenario characteristics, fault models, and energy consumption constraints to automatically generate optimal maintenance strategies. Summary of the Invention

[0006] This invention provides a method and system for recommending optimal maintenance solutions for the Internet of Things (IoT). It can automatically recommend maintenance modules and their parameter combinations under budget constraints by establishing an optimization model between overall maintenance cost and system reliability, taking into account factors such as network scale, node distribution, fault type and probability, and energy consumption model under different application scenarios. The method uses typical maintenance function modules such as remote restart, heartbeat detection, and firmware updates as candidate components that can be activated on demand. Combined with key parameters such as energy warning threshold and heartbeat frequency, it evaluates the actual effect of maintenance strategies based on system modeling and simulation analysis, thereby selecting the optimal maintenance solution. To support the quantitative comparison of the effectiveness of maintenance strategy configurations, a WSN maintainability index is proposed to evaluate the trade-off between system reliability and maintenance cost for different strategy combinations.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A method for recommending optimal maintenance solutions for the Internet of Things (IoT) includes:

[0009] Step 1: Obtain system input, which includes a set of maintenance modules, a set of fault models, and a set of WSN parameters. The set of maintenance modules contains multiple optional maintenance modules, each with a name, activation flag, development cycle, relevant parameters, and information on its impact on faults. The set of fault models contains various fault types, each with an occurrence probability and manual repair cost. The set of WSN parameters includes scenario design parameters, routing protocol parameters, energy consumption model parameters, cost model parameters, and maintenance rule parameters.

[0010] Step 2: Based on the system input, perform WSN simulation and cost assessment. The assessment includes simulating network operation status, fault evolution process and maintenance behavior response, and calculating the full life cycle cost. The full life cycle cost includes basic cost, maintenance module cost, periodic inspection cost, and fault and maintenance cost.

[0011] Step 3: Use an intelligent search algorithm to search for the optimal maintenance scheme under budget constraints. The intelligent search algorithm generates candidate maintenance configurations, inputs the candidate maintenance configurations into Step 2 for evaluation, determines whether the budget constraints are met, iteratively updates the optimal solution until the termination condition is met, and outputs the optimal maintenance scheme and corresponding cost. Among them, the candidate maintenance configuration includes the enabled combination of maintenance modules and the parameter configuration of the enabled modules, as well as the configuration of maintenance rule parameters in WSN parameters.

[0012] In this specification, in step 1, the maintenance modules in the maintenance module set are divided into a basic layer, a core layer, and a recovery layer; the basic layer includes a physical protection module, a communication assurance module, and a status monitoring module; the core layer includes an energy management module, a software management module, and an environmental adaptation module; the recovery layer includes a remote recovery module, a local recovery module, and a component recovery module; the impact information of each module includes at least one of fault prevention, reducing the probability of faults, and fault repair capability.

[0013] In this specification, in step 1, the fault types in the fault model set include communication faults, data loss, hardware faults, firmware faults, energy faults, data errors, and hidden node faults; the occurrence probability of each fault type is the basic fault probability. When a maintenance module that reduces the impact of the fault probability is enabled, the occurrence probability of the fault type is the basic fault probability multiplied by (1 - mitigation coefficient), where the mitigation coefficient is a relevant parameter of the maintenance module.

[0014] In this specification, in step 1, the scenario design parameters in the WSN parameter set include the number of sensor nodes, their distribution, coverage area, expected system lifespan, sampling frequency, and data size; the routing protocol parameters include forwarding rules, cluster formation, and routing selection indicators; the energy consumption model parameters include energy consumption per bit of transmission, reception, amplification, sampling, data aggregation, and initial battery energy; the cost model parameters include hardware price, development and installation costs, and project budget; and the maintenance rule parameters include inspection intervals and warning energy levels.

[0015] In this specification, step 2, the basic cost includes hardware procurement cost, system development cost, and installation and deployment cost; the maintenance module cost is the sum of the development cost and hardware cost of all activated maintenance modules. The development cost is determined by the module development cycle and personnel salaries, and the hardware cost is the procurement cost of additional hardware required by the module; the periodic inspection cost is the product of the number of inspections and the cost of each inspection. The number of inspections is determined by the ratio of the expected system lifespan to the inspection interval; the fault and maintenance costs include manual maintenance costs and data loss costs. The manual maintenance cost is the cost of manual repair when adaptive maintenance fails after a fault occurs, and the data loss cost is calculated from the value of the data packet.

[0016] In this specification, step 3, the process of constructing the candidate maintenance configuration is as follows: First, an activation combination is generated in the maintenance module space using an intelligent search strategy, and the activation flag of each module is determined; then, relevant parameters are configured for the activated maintenance modules, including heartbeat frequency, wireless charging power, etc.; finally, maintenance rule parameters in the WSN parameters are configured, including periodic inspection interval and remaining energy threshold.

[0017] In this specification, step 3, the intelligent search algorithm includes a genetic algorithm, a particle swarm optimization algorithm, or a simulated annealing algorithm; the genetic algorithm iteratively evolves candidate maintenance configurations through simulated natural selection, crossover, and mutation; the particle swarm optimization algorithm treats candidate configurations as particles and iteratively updates candidate configurations through information sharing among particles; the simulated annealing algorithm accepts better solutions probabilistically according to a temperature plan to avoid getting trapped in local optima.

[0018] In this specification, step 2, when performing WSN simulation and cost assessment, also includes evaluating the maintenance strategy based on maintainability indicators. The maintainability indicators are used to quantitatively measure the trade-off effect of the maintenance strategy between system reliability and maintenance cost. System reliability comprehensively considers the impact of the number of maintenance modules activated and the configuration of key parameters on reliability.

[0019] In this specification, in step 3, the iterative update process of the intelligent search is as follows: if the total lifecycle cost of the candidate maintenance configuration is lower than the current optimal solution and meets the budget constraint, then the optimal solution is updated; if the budget constraint is not met, then the candidate maintenance configuration is discarded; when the number of iterations reaches a preset value or the optimal solution is not updated for multiple consecutive times, the termination condition is met.

[0020] An optimal maintenance scheme recommendation system for the Internet of Things (IoT) applies any one of the above-mentioned optimal maintenance scheme recommendation methods. The optimal maintenance scheme recommendation system for the IoT includes:

[0021] The acquisition module is used to acquire system input, which includes a maintenance module set, a fault model set, and a WSN parameter set. The maintenance module set contains multiple optional maintenance modules, each with a name, activation flag, development cycle, relevant parameters, and information on its impact on faults. The fault model set contains various fault types, each with an occurrence probability and manual repair cost. The WSN parameter set includes scenario design parameters, routing protocol parameters, energy consumption model parameters, cost model parameters, and maintenance rule parameters.

[0022] The evaluation module is used to perform WSN simulation and cost evaluation based on the system input. The evaluation includes simulating network operation status, fault evolution process and maintenance behavior response, and calculating the full life cycle cost. The full life cycle cost includes basic cost, maintenance module cost, periodic inspection cost and fault and maintenance cost.

[0023] The search module is used to search for the optimal maintenance scheme under budget constraints using an intelligent search algorithm. The intelligent search algorithm generates candidate maintenance configurations, inputs the candidate maintenance configurations to the evaluation layer for evaluation, determines whether the budget constraints are met, iteratively updates the optimal solution until the termination condition is met, and outputs the optimal maintenance scheme and corresponding cost. The candidate maintenance configuration includes the enabled combination of maintenance modules and the parameter configuration of the enabled modules, as well as the configuration of maintenance rule parameters in the WSN parameters.

[0024] Unlike existing maintenance strategies in IoT or WSN systems that rely on experience-based configuration, static design, or heavy reliance on manual judgment, the optimal maintenance scheme recommendation method and system proposed in this invention have the following beneficial effects:

[0025] 1) Improve the scientific nature and relevance of maintenance strategy configuration: By constructing a quantifiable maintenance module model and maintenance cost calculation system, the traditional practice of relying on manual experience for module activation and parameter setting is eliminated, making the maintenance strategy configuration process more systematic and able to adaptively adjust to different deployment scenarios, significantly improving the adaptability and effectiveness of maintenance strategies.

[0026] 2) Significantly reduce maintenance costs and system downtime risk: By introducing multiple factors such as maintenance overhead, failure probability, and coverage performance to construct a comprehensive cost function, this invention can automatically recommend the maintenance module combination and parameter configuration with the lowest total cost while meeting the budget, avoiding the problems of over-maintenance and under-maintenance, thereby reducing unnecessary energy consumption and operation and maintenance costs, and improving the long-term stability of the system.

[0027] 3) Excellent scalability and applicability: The modular modeling approach allows for flexible combination and expansion of maintenance modules, adapting to various WSN deployment scenarios (such as smart meter reading, parking lot monitoring, and animal room monitoring), while supporting different routing protocols, and possessing high versatility and engineering migration capabilities.

[0028] 4) Applicable to resource-constrained edge IoT systems: Unlike predictive maintenance methods that rely on real-time monitoring and edge-cloud collaborative computing, this invention generates maintenance plans before system operation by modeling and optimizing solutions during the design phase. This avoids dependence on high computing power and continuous connectivity, and is especially suitable for low-power, low-computing, and highly distributed wireless sensor network systems. Attached Figure Description

[0029] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a schematic diagram of the method for recommending the optimal maintenance scheme for the Internet of Things involved in this invention.

[0031] Figure 2 This is a flowchart illustrating the method for recommending the optimal maintenance scheme for the Internet of Things involved in this invention.

[0032] Figure 3 This is a schematic diagram of the architecture of the wireless sensor network maintenance module involved in this invention. Detailed Implementation

[0033] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the embodiments of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0034] The following disclosure provides many different implementations or examples for carrying out different structures of the embodiments of the present invention. To simplify the disclosure of the embodiments of the present invention, specific examples of components and arrangements are described below. Of course, these are merely examples and are not intended to limit the embodiments of the present invention. Furthermore, reference numerals and / or reference letters may be repeated in different examples of the embodiments of the present invention; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various implementations and / or arrangements discussed.

[0035] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0036] like Figure 1 As shown, this embodiment provides a method for recommending the optimal maintenance scheme for the Internet of Things, including:

[0037] Step 1: Obtain system input, which includes a set of maintenance modules, a set of fault models, and a set of WSN parameters. The set of maintenance modules contains multiple optional maintenance modules, each with a name, activation flag, development cycle, relevant parameters, and information on its impact on faults. The set of fault models contains various fault types, each with an occurrence probability and manual repair cost. The set of WSN parameters includes scenario design parameters, routing protocol parameters, energy consumption model parameters, cost model parameters, and maintenance rule parameters.

[0038] Step 2: Based on the system input, perform WSN simulation and cost assessment. The assessment includes simulating network operation status, fault evolution process and maintenance behavior response, and calculating the full life cycle cost. The full life cycle cost includes basic cost, maintenance module cost, periodic inspection cost, and fault and maintenance cost.

[0039] Step 3: Use an intelligent search algorithm to search for the optimal maintenance scheme under budget constraints. The intelligent search algorithm generates candidate maintenance configurations, inputs the candidate maintenance configurations into Step 2 for evaluation, determines whether the budget constraints are met, iteratively updates the optimal solution until the termination condition is met, and outputs the optimal maintenance scheme and corresponding cost. Among them, the candidate maintenance configuration includes the enabled combination of maintenance modules and the parameter configuration of the enabled modules, as well as the configuration of maintenance rule parameters in WSN parameters.

[0040] In some embodiments, in step 1, the maintenance modules in the maintenance module set are divided into a basic layer, a core layer, and a recovery layer. The basic layer includes a physical protection module, a communication assurance module, and a status monitoring module. The core layer includes an energy management module, a software management module, and an environmental adaptation module. The recovery layer includes a remote recovery module, a local recovery module, and a component recovery module. The impact information of each module includes at least one of fault prevention, fault probability reduction, and fault repair capability. The basic layer modules mainly target low-level faults such as physical damage and communication conflicts; the core layer modules mainly mitigate continuous operation faults such as energy depletion and software defects; and the recovery layer modules mainly address hardware / firmware failures that have already occurred.

[0041] In some embodiments, in step 1, the fault types in the fault model set include communication faults, data loss, hardware faults, firmware faults, energy faults, data errors, and hidden node faults. The occurrence probability of each fault type is the basic fault probability. When a maintenance module that reduces the impact of fault probability is enabled, the occurrence probability of that fault type is the basic fault probability multiplied by (1 - mitigation coefficient), whereby the mitigation coefficient is a relevant parameter of the maintenance module. The mitigation coefficient is pre-calibrated based on historical fault data or experimental tests, or obtained by training a machine learning model (such as a random forest) on the fault suppression effect in similar scenarios, and is stored in the parameter library of the maintenance module.

[0042] In some embodiments, in step 1, the scenario design parameters in the WSN parameter set include the number, distribution, coverage area, expected system lifetime, sampling frequency, and data size of sensor nodes; the routing protocol parameters include forwarding rules, cluster formation, and routing selection metrics; the energy consumption model parameters include energy consumption per bit of transmission, reception, amplification, sampling, data aggregation, and initial battery energy; the cost model parameters include hardware price, development and installation costs, and project budget; and the maintenance rule parameters include check intervals and warning energy levels. The warning energy level in the maintenance rule parameters is used to trigger the activation threshold of the energy management module (such as wireless charging). When the remaining energy of a node is lower than this level, the wireless charging module is automatically activated.

[0043] In some embodiments, in step 2, the basic cost includes hardware procurement cost, system development cost, and installation and deployment cost; the maintenance module cost is the sum of the development cost and hardware cost of all enabled maintenance modules, with the development cost determined by the module development cycle and personnel salaries, and the hardware cost being the procurement cost of additional hardware required by the module; the periodic inspection cost is the product of the number of inspections and the cost of each inspection, with the number of inspections determined by the ratio of the expected system lifespan to the inspection interval; the fault and maintenance costs include manual maintenance costs and data loss costs, with manual maintenance costs being the cost of manual repair when adaptive maintenance fails after a fault occurs, and the data loss cost being calculated from the data packet value. The data packet value is preset according to its application scenario priority (e.g., industrial control data is higher than environmental monitoring data), or configured through the data importance weight table in the scenario parameter θ, which is input by the user according to business needs.

[0044] In some embodiments, the process of constructing the candidate maintenance configuration in step 3 is as follows: First, an enabling combination is generated in the maintenance module space using an intelligent search strategy, and the enabling flag of each module is determined; then, relevant parameters are configured for the enabled maintenance modules, including heartbeat frequency, wireless charging power, etc.; finally, maintenance rule parameters in the WSN parameters are configured, including periodic inspection interval and remaining energy threshold. Pre-screening is performed based on the correlation between maintenance modules and fault types (e.g., only modules that affect high-frequency faults are retained), and then the screened modules are combined and enumerated, or candidate combinations are generated using a binary encoding method through an intelligent algorithm.

[0045] In some embodiments, in step 3, the intelligent search algorithm includes a genetic algorithm, a particle swarm optimization algorithm, or a simulated annealing algorithm; the genetic algorithm iteratively evolves candidate maintenance configurations by simulating natural selection, crossover, and mutation; the particle swarm optimization algorithm treats candidate configurations as particles and iteratively updates candidate configurations through information sharing among particles; the simulated annealing algorithm accepts better solutions probabilistically according to a temperature plan to avoid getting trapped in local optima.

[0046] In some embodiments, step 2, during WSN simulation and cost assessment, further includes evaluating the maintenance strategy based on maintainability metrics. These maintainability metrics are used to quantitatively measure the trade-off between system reliability and maintenance cost. System reliability comprehensively considers the impact of the number of activated maintenance modules and the configuration of key parameters on reliability. In WSN simulation and cost assessment, maintainability metrics serve as an auxiliary optimization objective for intelligent search (forming a multi-objective function together with total cost), or as a screening threshold for feasible solutions (e.g., only retaining candidate configurations with maintainability metrics higher than a preset threshold).

[0047] In some embodiments, in step 3, the iterative update process of the intelligent search is as follows: if the total lifecycle cost of the candidate maintenance configuration is lower than the current optimal solution and meets the budget constraint, then the optimal solution is updated; if the budget constraint is not met, then the candidate maintenance configuration is discarded; when the number of iterations reaches a preset value or the optimal solution has not been updated for multiple consecutive iterations, the termination condition is met. The preset number of iterations is dynamically adjusted according to the WSN node scale (e.g., 50 times for less than 100 nodes, and 100 times for more than 500 nodes), and the threshold for the number of consecutive times the optimal solution has not been updated is 20% of the number of iterations.

[0048] In some embodiments, the optimal maintenance scheme output in step 3 includes the names of the enabled maintenance modules, the parameter values ​​of each enabled module, and the configuration values ​​of the maintenance rule parameters in the WSN parameters; the corresponding costs include the specific values ​​of the basic cost, maintenance module cost, periodic inspection cost, and fault and maintenance costs.

[0049] In some embodiments, during WSN operation, fault probabilities in the fault model are dynamically updated using a sliding window algorithm (e.g., data from the last 30 days) based on fault records (such as fault occurrence time and environmental parameters) uploaded in real time by edge nodes. The intelligent search module periodically (e.g., quarterly) re-optimizes the maintenance plan based on the updated fault model to adapt to the temporal changes in fault modes. The innovation lies in the fact that existing solutions rely on preset static fault probabilities, while in reality, fault probabilities dynamically change with environmental aging and node load variations. This mechanism enables the solution to be adaptive, a method unconventional in this field.

[0050] In some embodiments, during intelligent search, the search mode is automatically switched based on the complexity of the candidate solution space (e.g., high complexity is determined when the number of modules > 10 or the number of fault types > 5): Particle swarm optimization (fast convergence) is used for low-complexity scenarios, while a fusion algorithm of genetic algorithm and simulated annealing (genetic algorithm global search + simulated annealing local fine-tuning) is used for high-complexity scenarios. Furthermore, an algorithm performance evaluation model is trained using historical search data, and algorithm parameters (e.g., the crossover probability of the genetic algorithm) are dynamically adjusted. The innovation lies in the fact that existing solutions use a single algorithm and do not consider the impact of scenario complexity on algorithm efficiency. This strategy improves search accuracy and speed through scenario adaptation and algorithm fusion, demonstrating non-obviousness.

[0051] In some embodiments, when evaluating candidate maintenance configurations, a synergy coefficient matrix is ​​introduced to quantify the joint impact of any two enabled modules: if module A and module B have a cumulative effect on suppressing the same fault (e.g., the suppression rate of energy faults by wireless charging + remote restart is greater than the sum of their individual effects), the synergy coefficient is set to 1.2; if there is a canceling effect (e.g., high-frequency heartbeat detection + wireless charging leads to a surge in energy consumption), the synergy coefficient is set to 0.8; and the synergy coefficient is incorporated into the fault probability calculation and cost assessment. Creativity: Existing solutions evaluate the impact of individual modules in isolation, ignoring the interaction between modules. This model improves the accuracy of configuration optimization by quantifying synergistic effects, representing a novel technical concept.

[0052] In some embodiments, WSN simulation and cost assessment adopt a layered execution mode: edge nodes run simplified simulations locally (calculating only fault probabilities and energy consumption relevant to themselves) and upload the local results to the cloud; the cloud aggregates edge data and performs global simulations (such as the impact of routing protocols on the overall network and cross-node fault propagation); intelligent search is completed in the cloud, but the edge module parameters (such as heartbeat frequency) in the candidate configurations are fine-tuned by the edge nodes based on local resources (remaining power). The innovation lies in the fact that existing solutions rely on centralized simulation in the cloud, which is not suitable for resource-constrained edge nodes. This mechanism reduces the computational load on edge nodes through layered simulation while retaining global optimization capabilities, making it suitable for IoT edge computing scenarios.

[0053] An optimal maintenance scheme recommendation system for the Internet of Things (IoT) applies any one of the above-mentioned optimal maintenance scheme recommendation methods. The optimal maintenance scheme recommendation system for the IoT includes:

[0054] The acquisition module is used to acquire system input, which includes a maintenance module set, a fault model set, and a WSN parameter set. The maintenance module set contains multiple optional maintenance modules, each with a name, activation flag, development cycle, relevant parameters, and information on its impact on faults. The fault model set contains various fault types, each with an occurrence probability and manual repair cost. The WSN parameter set includes scenario design parameters, routing protocol parameters, energy consumption model parameters, cost model parameters, and maintenance rule parameters.

[0055] The evaluation module is used to perform WSN simulation and cost evaluation based on the system input. The evaluation includes simulating network operation status, fault evolution process and maintenance behavior response, and calculating the full life cycle cost. The full life cycle cost includes basic cost, maintenance module cost, periodic inspection cost and fault and maintenance cost.

[0056] The search module is used to search for the optimal maintenance scheme under budget constraints using an intelligent search algorithm. The intelligent search algorithm generates candidate maintenance configurations, inputs the candidate maintenance configurations to the evaluation layer for evaluation, determines whether the budget constraints are met, iteratively updates the optimal solution until the termination condition is met, and outputs the optimal maintenance scheme and corresponding cost. The candidate maintenance configuration includes the enabled combination of maintenance modules and the parameter configuration of the enabled modules, as well as the configuration of maintenance rule parameters in the WSN parameters.

[0057] The technical concept of this invention is as follows:

[0058] The recommended method and system for optimal maintenance solutions for the Internet of Things (IoT) achieves rational configuration of maintenance modules, quantitative estimation of fault impact, simulation of system behavior, and automatic search for the globally optimal strategy by constructing a standardized modeling and evaluation mechanism and composable maintenance strategies. Figure 2 The overall flowchart includes three key steps: system input, WSN simulation and cost assessment, and intelligent search. First, the system requires input including the maintenance module M, the fault model F, and the WSN parameters θ. The intelligent search algorithm then searches for the optimal combination of maintenance module and WSN parameters within budget constraints, minimizing the total cost during WSN simulation and cost assessment.

[0059] 1. Maintainability Definition

[0060] Maintainability refers to the system's ability to maintain high operational reliability while reducing failures and maintenance costs through the proper configuration of maintenance modules and parameters. Its calculation formula is as follows:

[0061]

[0062] Where μ∈(0,1) is the weight of reliability in maintainability calculation, and R∈[0,1] represents the system reliability. C failure C represents the cost of maintenance failures and data loss during the expected lifespan. total This represents the total system cost over its expected lifespan. The system reliability score R takes into account the enabled maintenance modules and key parameter configurations, and is calculated using the following formula:

[0063]

[0064] Where ω∈(0,1) is the weight of the proportion of activated maintenance modules to the total number of maintenance modules, and n moudle n represents the number of maintenance modules that are enabled. max This indicates the total number of maintenance modules. This represents the score of all key parameters in reliability. Among them, Normalize different parameters, s i Let r be the reliability score of the i-th parameter, and k be the total number of critical parameters that need maintenance and optimization. Critical parameters include global parameters and maintenance module parameters; therefore, when this parameter is a global parameter, r... i =1; When this parameter is a module parameter, if the module is enabled, then r i =1, otherwise r i =0. For some parameters, larger values ​​generally indicate higher reliability, while for others the opposite. Among these, para... i This represents the value of the i-th parameter. This is the minimum possible value for this parameter. This represents the maximum value that the parameter can take. Therefore, its formula is as follows:

[0065]

[0066] Therefore, maintainability, as an indicator for evaluating system maintainability, is used to quantitatively measure the overall effect of maintenance strategy configuration in terms of reliability improvement and cost control.

[0067] 2.1 System Input

[0068] In this stage, the system defines optional maintenance modules (such as RTS / CTS mechanism, remote restart, wireless charging module, etc.), typical fault types (such as communication failure, data loss, hardware failure, etc.) and their probabilities and manual repair costs in a modular manner, and configures WSN parameters (such as scenario design parameters, routing protocol parameters, etc.) to construct the input solution space required for intelligent search, WSN simulation and cost assessment.

[0069] 2.1.1 Maintenance Module M

[0070] 1) Definition

[0071] Definition 1: M represents the set of n optional maintenance modules, i.e., M = {m1, m2, ..., mn} n Each module m i It is a structured object: m i =(name i ,enabled i ,cycle i ,params i impact i ).

[0072] Wherein, `name` is the module name (e.g., heartbeat monitoring, wireless charging, etc.); `enabled` is a boolean flag indicating whether the maintenance module is selected; `cycle` represents the additional development time (in months) required to integrate the module into the WSN system; `params` encapsulates module-related parameters (e.g., heartbeat frequency, additional hardware costs, etc.); and `impact` represents the maintenance module's impact on mitigating specific fault types, which is divided into three main categories:

[0073] Fault prevention: This subset represents module m i A subset of failures that can be completely prevented due to their architectural design or inherent physical limitations. When failure f j In m i When the prevention set is defined, it indicates that when module m i When enabled, its fault f j The probability of it occurring is 0.

[0074] Reduce the probability of failure: indicates module m i Reduce failures j The ability to determine the probability of occurrence. If the fault type f... j The basic failure probability is p j The mitigation coefficient α defined in this module ij , indicating that when module m i When enabled, a fault f occurred. j The probability decreases, and the reduced probability is (1-α). ij )*p j .

[0075] Fault repair capability: indicates module m i It can capture fault f j The ability to occur and repair. Under this influence, the module defines the expected repair success rate α. ij That is, fault f j Once it happens, the probability of successful repair can be increased.

[0076] 2) Functional analysis of the maintenance module

[0077] Figure 3 This demonstrates a comprehensive maintenance module architecture designed to improve the maintainability of wireless sensor nodes. The architecture employs a layered design approach, dividing maintenance functions into a basic protection and monitoring layer, a core active maintenance layer, and a fault maintenance strategy layer.

[0078] First, the foundational layer forms the cornerstone for stable node operation. The physical protection layer (peripheral hardware modules) proactively resists potential harsh environments (such as humidity and impact) through enhanced shell design, reducing the risk of physical damage. The communication assurance layer (RTS / CTS mechanism) utilizes handshake protocols at the link layer to effectively avoid hidden terminal conflicts and ensure reliable data transmission. The status monitoring layer uses a heartbeat module as the system's "pulse," periodically reporting node status (including key parameters such as remaining battery power), providing real-time data for upper-layer maintenance decisions. Its configurability (heartbeat frequency, packet loss threshold) balances monitoring accuracy with communication overhead.

[0079] The core layer focuses on proactively maintaining nodes in optimal condition. The energy management zone primarily uses wireless charging modules to provide critical replenishment via contactless energy transfer when nodes face energy depletion, making it particularly suitable for harsh or inaccessible environments and directly combating power failures. The software management zone uses a firmware update module to support the remote deployment of verified software updates and patches, fundamentally preventing system failures caused by firmware defects or version incompatibility. The environmental adaptation zone (noise recalibration) dynamically senses and compensates for environmental electromagnetic interference, significantly improving the integrity and accuracy of data acquisition and reducing data loss due to environmental factors by adaptively adjusting internal filtering parameters.

[0080] The recovery layer provides a tiered fault response mechanism. The remote recovery zone (remote reboot / factory reset) is the preferred solution, supporting remote resolution of transient software errors (such as memory failures) or persistent configuration / data corruption issues via network, minimizing manual intervention. When the remote channel fails, the local recovery zone (local reboot / factory reset) provides a physical contact backup solution, ensuring the restoration of basic node functions even in high-interference or communication interruption scenarios. The component recovery zone (sensor reactivation) specifically addresses situations where sensor components enter a dormant or faulty state, restoring their sensing and communication capabilities through reinitialization.

[0081] Figure 3 The maintenance module architecture shown must be selectively configured according to strict cost-benefit trade-offs in actual deployment. Each module and its corresponding parameters differ in resource overhead (energy consumption, hardware costs, etc.) and fault suppression performance, requiring optimization to achieve the best combination.

[0082] 2.1.2 Fault Model F

[0083] 1) Definition

[0084] Definition 2: F = {(f1, p1, c1), ..., (f k ,p k ,c k )} indicates that there are a total of k fault models. Where f j Indicates the fault type, p j c represents the corresponding basic failure probability. j This indicates the cost of manually repairing this type of fault.

[0085] 2) Fault type analysis

[0086] Wireless sensor networks (WSNs) operate in complex, dynamic, and resource-constrained environments. Their nodes are susceptible to various internal and external factors, leading to functional abnormalities or failures, collectively referred to as faults. These faults not only impair the performance of individual nodes but also create cascading effects at the network level through inter-node cooperation, potentially impacting overall network performance, data quality, and lifespan. Based on in-depth analysis of typical WSN deployment scenarios, this model primarily considers the following seven types of critical faults.

[0087] Hidden Nodes: This problem stems from the classic hidden terminal problem. When two nodes located outside each other's wireless transmission range simultaneously attempt to communicate with a common receiving node, the lack of an effective carrier sensing mechanism causes their transmitted signals to collide at the receiving node, resulting in packet loss and unstable communication links.

[0088] Data errors: These manifest as incorrect, inconsistent, or logically corrupted data values ​​acquired or generated by the nodes. This is typically caused by environmental interference, long-term drift of the sensor itself, or calculation errors during node processing. Specific forms include abnormal readings exceeding the physical measurement range, structural errors violating predetermined data formats, or logical contradictions.

[0089] Data loss refers to the partial or complete absence of sensor data during data transmission or local storage at nodes. Unstable network connections, node buffer overflows, communication interruptions, or unexpected node restarts are the main causes. Data loss disrupts the continuity and integrity of data collection, leading to gaps in the information flow.

[0090] Communication failure manifests as a node's inability to establish or maintain a reliable information exchange channel with its neighboring nodes or the target receiving node. The causes of this failure are varied, including excessive interference to the wireless channel, degradation of antenna physical performance, hardware failure of the wireless transceiver module, and misconfiguration of the network protocol stack. Persistent communication failures can isolate nodes, even leading to network partitioning, disrupting network connectivity, and hindering information convergence.

[0091] Energy failure refers to a node's complete loss of operational capability due to the depletion of its primary energy resources. The root causes include initial battery capacity limitations, the node being in a high-energy-consumption state for extended periods, or a lack of effective energy replenishment mechanisms. Energy failure directly leads to node "death," irreversibly reducing network size and shortening the overall network lifespan.

[0092] Hardware failure refers to physical damage or malfunction of internal components of the sensor node, including the microcontroller, memory, sensor, or wireless transceiver. Such failures can be caused by harsh environmental conditions, mechanical shock, or manufacturing defects.

[0093] Firmware failure: This is a software-level problem, which manifests as the node failing to enter normal working state during power-on initialization or system reset because it cannot correctly load or execute its firmware program.

[0094] 2.1.3 WSN parameter θ

[0095] θ represents system-level configuration parameters, including scene setting parameters θ scenario Routing protocol parameters θ routing Energy consumption model parameter θ energy Cost model parameter θ cost and maintaining rule parameters θ rule These parameters collectively influence the system's behavior and cost assessment, and are subject to specific deployment constraints.

[0096] Scene parameter setting parameter θ scenario This defines the scene parameters of the WSN, such as the number and distribution of sensor nodes, the coverage area of ​​the scene, the expected system lifespan, sampling frequency, and data size.

[0097] Routing protocol parameter θ routing This describes the internal behavior of the data transmission protocol, such as forwarding rules, cluster formation, and routing metrics. It does this by exposing an abstract interface for protocol parameters (e.g., parameters required by the LEARN MEL protocol include: cluster head ratio, maximum hop count, and number of frames per round).

[0098] Energy consumption model parameter θ energy This refers to the energy consumption characteristics of sensor hardware, including energy consumption per bit for transmission, reception, amplification, sampling, data aggregation, and initial battery power.

[0099] Cost model parameter θ cost This is used to calculate development, operating, and maintenance costs. It includes hardware prices, development and installation fees, and the overall project budget. These inputs are essential for assessing lifecycle costs.

[0100] Maintain rule parameter θ rule: Indicates the rules and thresholds that need to be configured for preventative maintenance strategies, such as check intervals and warning energy levels. These parameters affect when and how interventions are triggered.

[0101] 2.2 WSN Simulation and Cost Assessment

[0102] The system runs a simulation model based on the input WSN combination configuration, dynamically simulating network operation status, fault evolution process, and maintenance response. It evaluates the cost of each maintenance strategy through an integrated cost model. To quantitatively assess the economic impact of different maintenance configurations in the WSN, we define a unified cost model. The total lifecycle cost integrates basic costs, module-related costs, periodic inspection costs, losses caused by faults, and maintenance costs. Each cost component is influenced by a specific subset of parameters from the maintenance module M, fault model F, or configuration parameters θ.

[0103] 1) Basic cost C fixed This includes all one-time expenses related to hardware procurement, system development, and on-site installation and deployment. These scenario values ​​are determined by θ. scenario and θ cost It was decided.

[0104] C fixed =C hardware +C development +C installation ;

[0105] Among them, C hardware C represents the hardware cost of sensor nodes and base stations. decelopment For the development cost of the WSN system, C installation The deployment cost of the WSN system.

[0106] 2) Maintenance module cost C moudle Each selected maintenance module incurs corresponding development or hardware costs. These costs originate from the maintenance module m. i The module-specific parameters (cycle and params) in the code.

[0107] C moudle =∑m i ∈M,enabled i =1[C dev (m i )+C hw (m i )];

[0108] Among them, C dev (m i ) is module m i The development cost is determined by the extended development time and personnel salaries, while Chw (m i ) represents the additional hardware cost required to use this module, such as the need to purchase peripheral equipment for some modules.

[0109] 3) Cost of regular inspections C inspection According to θ rule The defined maintenance rules will trigger preventative fault checks periodically. The total number of checks depends on the scenario parameter θ. scenario Expected system lifetime and θ rule The check interval configured in the settings.

[0110]

[0111] Among them, T lifetime θ represents the scene parameters scenario Expected system lifetime, T interval Represents θ rule The check interval configured in the settings.

[0112] The total inspection cost is shown below:

[0113] C inspection =N check ×C check ;

[0114] Among them, C check Represents θ rule The cost per check configured in the settings.

[0115] 4) Failure and maintenance costs C failure During WSN operation, fault events trigger corresponding maintenance responses. These responses and their associated costs depend on the selected maintenance modules. This cost component includes energy consumption during adaptive maintenance, manual maintenance costs, and data loss costs. In the WSN simulation evaluation process, different fault types are first simulated based on the fault model, and the selected modules and parameters are used to determine whether the fault can be completely prevented or its probability reduced. Secondly, when a fault occurs, adaptive maintenance is performed based on the selected maintenance modules and parameters, which consumes energy. If adaptive maintenance fails, manual maintenance is performed, incurring manual maintenance costs C. repair When node failures or malfunctions are not detected in a timely manner, packet loss can occur, resulting in data loss cost C. loss , by θ scenario The corresponding data packet value is calculated.

[0116] C failure =C repair +C loss ;

[0117] 5) Expected total lifecycle cost Ctotal as follows:

[0118] C total =C fixed +C moudle +C inspection +C failure ;

[0119] 2.3 Intelligent Search

[0120] To achieve cost-effective proactive maintenance strategy configuration in resource-constrained WSN systems, this invention combines maintenance module modeling, fault feature analysis, and intelligent algorithm search to automatically generate and filter the optimal combination of maintenance modules and their parameter configurations.

[0121] like Figure 2 As shown, the entire intelligent search process iteratively executes four steps: "generating candidate configurations – WSN simulation and cost assessment – ​​budget constraint judgment – ​​optimal solution update," until the termination condition is met, at which point the optimal maintenance plan and corresponding cost are output. This process integrates the search capabilities of the algorithm with the feedback mechanism of the simulation system, possessing scalability to adapt to various application scenarios and fault types.

[0122] 2.3.1 Construction of Candidate Solution Space

[0123] The candidate solution space (i.e., candidate configuration) for the maintenance strategy consists of two parts: maintenance module combinations and tunable parameter configurations. The system is based on a maintenance module library, fault models, and WSN parameters. First, an intelligent search strategy is used to generate combinations in the maintenance module space, determining whether each module is enabled. Then, tunable operating parameters, such as heartbeat frequency and heartbeat packet loss threshold, are configured for the enabled modules. Finally, tunable maintenance rule parameters from the WSN parameters, such as periodic inspection intervals and remaining energy thresholds, are configured. This "combination + parameter" structure determines the preventive and repair capabilities of the maintenance strategy.

[0124] Due to the large number of available modules and the continuous nature of their parameters, the candidate space is enormous, making an exhaustive search insufficient for effective optimization. Therefore, intelligent algorithms are needed to optimize within the mixed space of discrete and continuous variables.

[0125] 2.3.2 Optimization Problem Modeling

[0126] This problem can be formalized as a multivariate constrained optimization problem, with the goal of minimizing the overall cost of the maintenance strategy over the entire network lifecycle while satisfying budget constraints. The optimization model is as follows:

[0127] 1) Optimization Variables: The decision variables in this optimization problem consist of two tightly coupled components. The first is the maintenance module configuration M, where each module m... i∈M contains an enable flag indicating whether it is selected (enabled) i =1), and related adjustable parameters (params) i The second component is the system-level configuration θ, which includes maintenance rules θ. rule The check interval and battery threshold warnings are parameters that need to be optimized.

[0128] 2) Objective Function and Constraints: The optimization objective is to determine the module selection and parameter configuration that minimizes the overall lifecycle cost of the WSN system under specific scenarios and budget constraints. The objective formula is: The constraints are:

[0129] C fixed +C moudle (M)+C inspection (θ)≤B;

[0130] (M * ,θ * )∈ValidConfig(F);

[0131] Where B is the scene configuration θ scenario The budget, with constraint value ValidConfig(F), ensures that the configuration meets deployment feasibility and the given failure characteristics F. Total cost C total This includes basic costs, module costs, regular inspection costs, and costs related to failures and maintenance.

[0132] 2.3.3 Optimization Process of Simulation-Driven System

[0133] Driven by intelligent algorithms, the entire optimization process iterates continuously, gradually approaching the optimal maintenance strategy. Its core process is as follows:

[0134] (1) Generate candidate configurations: The algorithm generates new maintenance module combinations and parameter configurations according to the coding rules.

[0135] (2) Simulation evaluation: Input the candidate configuration into the maintenance simulation system, simulate its behavior within a predetermined operating cycle, and evaluate data such as fault impact, maintenance response and energy consumption.

[0136] (3) Feasibility assessment and cost evaluation: Determine whether the configuration meets the budget requirements and calculate its total life cycle cost.

[0137] (4) Optimal solution update and iteration control: If the current candidate is better than the existing optimal solution, it is replaced; if the termination condition is met, the optimal configuration is output; otherwise, the next round of iteration continues.

[0138] 2.3.4 Optimization Algorithm Implementation

[0139] To adapt to mixed variable structures and nonlinear objective functions, this invention introduces three typical metaheuristic algorithms for comparative testing:

[0140] Genetic Algorithm (GA): A population-based evolutionary optimization method that iteratively evolves candidate maintenance configurations by simulating natural selection, crossover, and mutation. It is well-suited for problems involving joint binary continuous decision spaces.

[0141] Particle Swarm Optimization (PSO) is a swarm intelligence technique based on social behavior that models particles as candidate solutions that learn from the optimal positions of themselves and other particles. PSO is particularly effective for continuous parameter tuning.

[0142] Simulated Annealing (SA): Inspired by thermodynamic annealing, it explores the solution space by probabilistically accepting worse solutions according to a temperature plan, thus enabling it to escape local optima.

[0143] 2.3.5 Optimal Maintenance Strategy Output and Cost Decomposition

[0144] After the algorithm terminates, the system outputs the optimal subset M of modules. * ∈M and the corresponding system-level configuration θ * M * For the maintenance modules that have been enabled and their corresponding adjusted parameter values, θ * For θ rule The inspection interval and power threshold warning values ​​are corresponding to the adjusted values. To better analyze the final optimized configuration results, we also output the expected total lifecycle cost obtained from the simulation of the optimal configuration, including basic cost, module-related cost, periodic inspection cost, losses caused by failure, and maintenance cost.

[0145] In summary, this invention constructs a maintenance strategy recommendation method and system for resource-constrained wireless sensor networks. Unlike traditional configuration methods that rely on manual experience, this method can automatically recommend maintenance module combinations and their parameter configurations during the system design phase, based on the fault distribution characteristics, node resource conditions, and task requirements of specific scenarios, thereby achieving a balance between maintenance costs and system reliability.

[0146] The embodiments described above are for illustrative purposes only and are not intended to limit the invention. Therefore, any changes in numerical values ​​or substitutions of equivalent elements should still fall within the scope of this invention.

[0147] The above detailed description will enable those skilled in the art to understand that the present invention can indeed achieve the aforementioned objectives and has complied with the provisions of the Patent Law.

[0148] 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 the preferred embodiments as well as all changes and modifications falling within the scope of the invention. The above descriptions are merely preferred embodiments of the invention and are not intended to limit the invention. It should be noted that any modifications, equivalent substitutions, and improvements made within the spirit and principles of the invention should be included within the scope of protection of the invention.

[0149] It should be noted that the above description of the process is for illustrative purposes only and does not limit the scope of this specification. Those skilled in the art can make various modifications and changes to the process under the guidance of this specification. However, these modifications and changes remain within the scope of this specification.

[0150] The basic concepts have been described above. Obviously, for those skilled in the art who have read this application, the above disclosure is merely illustrative and does not constitute a limitation of this application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this application. Such modifications, improvements, and corrections are suggested in this application, and therefore, such modifications, improvements, and corrections still fall within the spirit and scope of the exemplary embodiments of this application.

[0151] Furthermore, this application uses specific terms to describe its embodiments. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that "an embodiment," "one embodiment," or "an alternative embodiment" mentioned twice or more in different positions in this specification do not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this application can be appropriately combined.

[0152] Furthermore, those skilled in the art will understand that aspects of this application can be described and illustrated through several patentable types or situations, including any new and useful combination of processes, machines, products, or substances, or any new and useful improvements thereof. Therefore, aspects of this application can be implemented entirely in hardware, entirely in software (including firmware, resident software, microcode, etc.), or a combination of hardware and software. All of the above hardware or software can be referred to as a “unit,” “module,” or “system.” Furthermore, aspects of this application can take the form of a computer program product embodied in one or more computer-readable media, wherein computer-readable program code is contained therein.

[0153] The computer program code required for the operation of each part of this application can be written in any one or more programming languages, including object-oriented programming languages ​​such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, and Python; general programming languages ​​such as C; Visual Basic, Fortran2103, Perl, COBOL2102, PHP, and ABAP; dynamic programming languages ​​such as Python, Ruby, and Groovy; or other programming languages. This program code can run entirely on the user's computer, or as a standalone software package on the user's computer, or partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer via any network, such as a local area network (LAN) or wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or used as a service such as Software as a Service (SaaS).

[0154] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this application are not intended to limit the order of the processes and methods of this application. Although some currently considered useful embodiments of the invention have been discussed in the foregoing disclosure by way of various examples, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the substance and scope of the embodiments of this application. For example, although the implementation of the various components described above can be embodied in a hardware device, it can also be implemented as a purely software solution, such as an installation on an existing server or mobile device.

[0155] Similarly, it should be noted that, in order to simplify the description of the present application and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of the embodiments of the present application sometimes combines multiple features into a single embodiment, drawing, or description thereof. However, this approach of the present application should not be construed as reflecting an intention that the claimed subject matter requires more features than expressly recited in each claim. Rather, the subject of the invention should possess fewer features than in any single embodiment described above.

Claims

1. A method for recommending optimal maintenance solutions for the Internet of Things (IoT), characterized in that, include: Step 1: Obtain system input, which includes a set of maintenance modules, a set of fault models, and a set of WSN parameters. The set of maintenance modules contains multiple optional maintenance modules, each with a name, activation flag, development cycle, relevant parameters, and information on its impact on faults. The set of fault models contains various fault types, each with an occurrence probability and manual repair cost. The set of WSN parameters includes scenario design parameters, routing protocol parameters, energy consumption model parameters, cost model parameters, and maintenance rule parameters. Step 2: Based on the system input, perform WSN simulation and cost assessment. The assessment includes simulating network operation status, fault evolution process and maintenance behavior response, and calculating the full life cycle cost. The full life cycle cost includes basic cost, maintenance module cost, periodic inspection cost, and fault and maintenance cost. Step 3: Use an intelligent search algorithm to search for the optimal maintenance scheme under budget constraints. The intelligent search algorithm generates candidate maintenance configurations, inputs the candidate maintenance configurations into Step 2 for evaluation, determines whether the budget constraints are met, iteratively updates the optimal solution until the termination condition is met, and outputs the optimal maintenance scheme and corresponding cost. Among them, the candidate maintenance configuration includes the enabled combination of maintenance modules and the parameter configuration of the enabled modules, as well as the configuration of maintenance rule parameters in WSN parameters.

2. The method for recommending the optimal maintenance scheme for the Internet of Things according to claim 1, characterized in that, In step 1, the maintenance modules in the maintenance module set are divided into a basic layer, a core layer, and a recovery layer; the basic layer includes a physical protection module, a communication assurance module, and a status monitoring module; the core layer includes an energy management module, a software management module, and an environmental adaptation module; the recovery layer includes a remote recovery module, a short-range recovery module, and a component recovery module; the impact information of each module includes at least one of fault prevention, fault probability reduction, and fault repair capability.

3. The method for recommending the optimal maintenance scheme for the Internet of Things according to claim 1, characterized in that, In step 1, the fault types in the fault model set include communication faults, data loss, hardware faults, firmware faults, energy faults, data errors, and hidden node faults. The occurrence probability of each fault type is the basic fault probability. When a maintenance module that reduces the impact of the fault probability is enabled, the occurrence probability of the fault type is the basic fault probability multiplied by (1 - mitigation coefficient), where the mitigation coefficient is a relevant parameter of the maintenance module.

4. The method for recommending the optimal maintenance scheme for the Internet of Things according to claim 1, characterized in that, In step 1, the scenario design parameters in the WSN parameter set include the number of sensor nodes, their distribution, coverage area, expected system lifespan, sampling frequency, and data size; the routing protocol parameters include forwarding rules, cluster formation, and routing selection indicators; the energy consumption model parameters include energy consumption per bit of transmission, reception, amplification, sampling, data aggregation, and initial battery energy; the cost model parameters include hardware price, development and installation costs, and project budget; and the maintenance rule parameters include inspection intervals and warning energy levels.

5. The method for recommending the optimal maintenance scheme for the Internet of Things according to claim 1, characterized in that, In step 2, the basic costs include hardware procurement costs, system development costs, and installation and deployment costs; the maintenance module cost is the sum of the development costs and hardware costs of all activated maintenance modules. The development cost is determined by the module development cycle and personnel salaries, and the hardware cost is the procurement cost of additional hardware required by the module; the periodic inspection cost is the product of the number of inspections and the cost of each inspection. The number of inspections is determined by the ratio of the expected system lifespan to the inspection interval; the fault and maintenance costs include manual maintenance costs and data loss costs. The manual maintenance cost is the cost of manual repair when adaptive maintenance fails after a fault occurs, and the data loss cost is calculated from the value of the data packets.

6. The method for recommending the optimal maintenance scheme for the Internet of Things according to claim 1, characterized in that, In step 3, the process of constructing the candidate maintenance configuration is as follows: First, an activation combination is generated in the maintenance module space using an intelligent search strategy to determine the activation flag of each module; then, relevant parameters are configured for the activated maintenance modules, including heartbeat frequency and wireless charging power; finally, maintenance rule parameters in the WSN parameters are configured, including periodic inspection interval and remaining energy threshold.

7. The method for recommending the optimal maintenance scheme for the Internet of Things according to claim 1, characterized in that, In step 3, the intelligent search algorithm includes a genetic algorithm, a particle swarm optimization algorithm, or a simulated annealing algorithm; the genetic algorithm iteratively evolves candidate maintenance configurations through simulated natural selection, crossover, and mutation; the particle swarm optimization algorithm treats candidate configurations as particles and iteratively updates candidate configurations through information sharing among particles; the simulated annealing algorithm accepts better solutions probabilistically according to a temperature plan to avoid getting trapped in local optima.

8. The method for recommending the optimal maintenance scheme for the Internet of Things according to claim 1, characterized in that, In step 2, when performing WSN simulation and cost assessment, the maintenance strategy is evaluated based on maintainability indicators. These maintainability indicators are used to quantitatively measure the trade-off between system reliability and maintenance cost. System reliability comprehensively considers the impact of the number of maintenance modules activated and the configuration of key parameters on reliability.

9. The method for recommending the optimal maintenance scheme for the Internet of Things according to claim 1, characterized in that, In step 3, the iterative update process of the intelligent search is as follows: if the total lifecycle cost of the candidate maintenance configuration is lower than the current optimal solution and the budget constraint is met, then the optimal solution is updated; if the budget constraint is not met, then the candidate maintenance configuration is discarded; when the number of iterations reaches a preset value or the optimal solution is not updated for multiple consecutive times, the termination condition is met.

10. An optimal maintenance solution recommendation system for the Internet of Things, characterized in that, The IoT optimal maintenance scheme recommendation system, which applies the method for recommending optimal maintenance schemes for any one of claims 1 to 9, comprises: The acquisition module is used to acquire system input, which includes a maintenance module set, a fault model set, and a WSN parameter set. The maintenance module set contains multiple optional maintenance modules, each with a name, activation flag, development cycle, relevant parameters, and information on its impact on faults. The fault model set contains various fault types, each with an occurrence probability and manual repair cost. The WSN parameter set includes scenario design parameters, routing protocol parameters, energy consumption model parameters, cost model parameters, and maintenance rule parameters. The evaluation module is used to perform WSN simulation and cost evaluation based on the system input. The evaluation includes simulating network operation status, fault evolution process and maintenance behavior response, and calculating the full life cycle cost. The full life cycle cost includes basic cost, maintenance module cost, periodic inspection cost and fault and maintenance cost. The search module is used to search for the optimal maintenance scheme under budget constraints using an intelligent search algorithm. The intelligent search algorithm generates candidate maintenance configurations, inputs the candidate maintenance configurations to the evaluation layer for evaluation, determines whether the budget constraints are met, iteratively updates the optimal solution until the termination condition is met, and outputs the optimal maintenance scheme and corresponding cost. The candidate maintenance configuration includes the activation combination of maintenance modules and the parameter configuration of the activation modules, as well as the configuration of maintenance rule parameters in the WSN parameters.

Images