Collaborative network system control method and device for intelligent device, and collaborative network system

Abstract

The application relates to the technical field of smart home, and discloses a cooperative network system control method for a smart device, which comprises the following steps: collecting operation state data of the smart device, and preprocessing the collected operation state data; performing fault analysis on the preprocessed operation state data, and determining fault information in the case that the analysis shows that there is a fault; using a preset model to infer a fault information propagation path to determine a fault root; performing reverse rollback on a corresponding control instruction according to the fault root, and controlling the smart device to issue an adjustment instruction. The method can realize fault positioning and self-repairing, and improves the stability and reliability of cooperative operation of the device. The application further discloses a cooperative network system control device for the smart device and a cooperative network system.

Description

Technical Field

[0001] This application relates to the field of smart home technology, such as a control method and apparatus for a collaborative network system for smart devices, and a collaborative network system. Background Technology

[0002] With the widespread application of smart devices in various scenarios, the collaborative operation between these devices is becoming increasingly critical. In smart home and similar scenarios where devices operate collaboratively, traditional fault handling methods struggle to accurately determine the fault propagation path after a device malfunctions, leading to time-consuming troubleshooting and severely impacting the efficiency and stability of device collaboration. Particularly during periods of continuous failure, ensuring system security while maintaining basic functionality becomes impossible, potentially resulting in complete device paralysis or security risks. Furthermore, with the use and upgrading of devices, version management issues become prominent. After manufacturers cease maintenance, maintaining stable control of older devices becomes difficult, and version switching can lead to lengthy interruptions and configuration drift, degrading the user experience.

[0003] Current technologies, when smart devices malfunction, only provide users with simple error messages, lacking in-depth analysis and precise location of the fault propagation path. For example, when multiple devices fail in collaborative operation, it's impossible to pinpoint which device's failure triggered the chain reaction, requiring a piecemeal troubleshooting approach that is extremely inefficient. Therefore, to address the technical problems of ambiguous fault location and weak self-healing capabilities in collaborative device scenarios, there is an urgent need for a self-healing network control scheme for collaborative smart devices. This scheme should enable fault location and self-healing repair, improving the stability and reliability of collaborative device operation.

[0004] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or describe the scope of protection of these embodiments, but rather as a prelude to the detailed description that follows.

[0006] This disclosure provides a method and apparatus for controlling a collaborative network system for smart devices, as well as a collaborative network system, to achieve fault location and self-healing repair, thereby improving the stability and reliability of collaborative device operation.

[0007] In some embodiments, the method includes: collecting operating status data of a smart device and preprocessing the collected operating status data; performing fault analysis on the preprocessed operating status data, and determining fault information if a fault exists; using a preset model to infer the propagation path of the fault information to determine the root cause of the fault; and based on the root cause of the fault, performing reverse rollback on the corresponding control commands and controlling the smart device to issue adjustment commands.

[0008] In some embodiments, the apparatus includes a processor and a memory storing program instructions, the processor being configured to, when executing the program instructions, perform the aforementioned collaborative network system control method for smart devices.

[0009] In some embodiments, the collaborative network system includes: a data acquisition and preprocessing module configured to acquire operating status data of intelligent devices and preprocess the acquired operating status data; a fault analysis module configured to perform fault analysis on the preprocessed operating status data, and determine fault information if a fault is found; a causal reasoning module configured to use a preset model to reason about the propagation path of the fault information to determine the root cause of the fault; and a reverse rollback module configured to perform reverse rollback on relevant control commands based on the root cause of the fault, and control the intelligent devices to issue adjustment commands.

[0010] The collaborative network system control method and apparatus for smart devices and the collaborative network system provided in this disclosure can achieve the following technical effects: The system collects operational status data from intelligent devices, preprocesses it, and then performs fault analysis. Using a pre-defined model, it performs causal reasoning to accurately pinpoint the fault propagation path and root cause. Then, by reversing control commands and adjusting commands issued by the intelligent devices, it eliminates the fault triggers and regulates the state of the intelligent devices. This achieves fault isolation and system self-healing, improves fault handling efficiency, and ensures the normal operation and stability of collaborative work of the equipment.

[0011] The above general description and the description below are exemplary and illustrative only and are not intended to limit this application. Attached Figure Description

[0012] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0013] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a schematic diagram of a collaborative network system control method for smart devices provided in an embodiment of this disclosure; Figure 2 This is a schematic flowchart of the method for constructing a causal relationship model provided in the embodiments of this disclosure; Figure 3 This is a flowchart illustrating the method provided in this embodiment of the present disclosure for determining the root cause of a fault by using a causal relationship model to infer the propagation path of fault information. Figure 4 This is a schematic diagram of another collaborative network system control method for smart devices provided in this disclosure embodiment; Figure 5 This is a schematic diagram of a collaborative network system control device for smart devices provided in an embodiment of this disclosure; Figure 6 This is a schematic diagram of a collaborative network system for smart devices provided in an embodiment of this disclosure. Detailed Implementation

[0015] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0016] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0017] According to one aspect of the embodiments of this application, a method for controlling a collaborative network system for smart devices is provided. This method is widely applicable to digital control scenarios such as smart homes, smart home ecosystems, intelligence house ecosystems, whole-house intelligence, digital homes, networked homes, automated homes, electronic homes, digital homes, and digital homes. It relies on technologies such as the Internet of Things (IoT), cloud computing, artificial intelligence (AI), digital technologies, digital twins, and AR / VR to build an intelligent, automated, personalized, and customized digital ecosystem. Optionally, in this embodiment, the above method can be applied to a hardware environment including a server. Cloud computing and / or edge computing services can be configured on the server or independently of the server to provide data processing services to the server.

[0018] Smart devices include, but are not limited to, PCs, mobile phones, tablets, as well as all categories of digital home appliances, digital appliances, smart home appliances, and whole-house smart devices. Specifically, these include: smart air conditioners, air conditioning equipment, smart range hoods, smart refrigerators, smart ovens, grilling and cooking equipment, smart stoves, smart washing machines, washing, rinsing, drying and garment processing equipment, smart water heaters, smart washing equipment, smart dishwashers, smart projectors, smart TVs, smart clothes racks, smart curtains, smart audio-visual equipment, smart sockets, smart speakers, smart speakers, smart ventilation equipment, smart kitchen and bathroom systems, kitchens, cabinets, kitchen cabinet equipment, smart bathroom equipment, robot vacuum cleaners, robot window cleaners, robot mopping, air purifiers, smart steam ovens, smart microwave ovens, smart water heaters, smart air purifiers, smart water dispensers, smart door locks, smart lighting fixtures, smart light bulbs, smart air detectors, smart remote controls, and smart gateways. This also covers smart terminals for security, lighting, and other digital home accessories.

[0019] Combination Figure 1 As shown, this disclosure provides a collaborative network system control method for smart devices, including: S101, the processor collects the operating status data of the smart device and preprocesses the collected operating status data.

[0020] S102, the processor performs fault analysis on the preprocessed running status data, and if a fault is found, it determines the fault information.

[0021] S103, the processor uses a preset model to infer the propagation path of fault information to determine the root cause of the fault.

[0022] S104: Based on the root cause of the fault, the processor reverses and rolls back the corresponding control instructions, and controls the intelligent device to issue adjustment instructions.

[0023] Here, the operational status data of intelligent devices includes at least device performance parameters and operation sequence records. Device performance parameters include real-time operating parameters, environmental parameters, and status parameters. In some scenarios, operational status data also includes energy consumption data. Operation sequence records include basic operation information and parameter change data after operation execution. Basic operation information includes the time, type, and identifier of the operation. After collecting operational status data, the data is cleaned to remove noise and invalid data, thus standardizing the data. Key features are extracted from the cleaned operational status data to provide accurate data for subsequent analysis.

[0024] Fault detection and classification are based on preprocessed operational status data. Algorithms such as static threshold algorithms, dynamic threshold algorithms, trend analysis algorithms, and intelligent recognition algorithms, or combinations thereof, can be used to perform multi-dimensional detection on the preprocessed data. Specifically, the static threshold algorithm, based on device hardware parameters and operating specifications, presets a safe threshold range for key indicators. If the preprocessed data exceeds this range and persists for a certain duration, it is determined to be a fault. For example, a brightness threshold for a smart bulb can be set, with a normal operating range of 5% to 100%. If the brightness is ≤5% and persists for ≥2 seconds, it is determined to be a brightness abnormality fault.

[0025] The dynamic threshold algorithm is based on historical operating data and calculates a dynamic threshold range through a sliding window to adapt to changes in device operating status. For example, for smart lighting fixtures, if the average brightness fluctuation in night mode over the past 7 days is calculated to be 60% with a variance of 5%, then the dynamic threshold range is set to 50%~70%. If the lighting fixture's brightness data falls outside the set dynamic threshold range for multiple consecutive collection periods, it is considered abnormal.

[0026] Trend analysis algorithms, such as linear regression / exponential smoothing fitting, fit preprocessed time-series data to analyze data trends. If the trend slope exceeds the normal range, it is identified as a precursor to a fault. For example, linear regression fitting of the brightness data of smart lamps reveals a continuous decrease in brightness at a rate of 0.5% per hour. Combined with a filament aging attenuation model, this is identified as a latent filament fault. Another example is exponential smoothing of the signal strength of the control center of a smart device. If the signal strength shows a continuous downward trend, a communication fault warning is issued.

[0027] Intelligent identification algorithms include machine learning classification algorithms and anomaly detection algorithms. Among these, machine learning classification algorithms, such as the random forest algorithm, use historical fault data as a training set, inputting multi-dimensional features to train a fault identification model. In the detection phase, real-time data is input into the model, and the model outputs the fault probability. For example, for a fault identification model of a lighting system, if real-time input data for a lamp (brightness = 0, current = 0, voltage = 220V, signal strength = 90%) is given, and the model outputs a 98% probability of filament burnout, then it is determined to be a filament burnout fault.

[0028] After a fault is detected, it is categorized to determine the fault information. Fault categories include hardware faults, software faults, communication faults, and environment-triggered faults, among others. Hardware faults include sensor faults, motor faults, and heating wire faults; software faults include program logic errors and version compatibility faults; communication faults include signal interference and link interruptions; and environment-triggered faults include faults caused by exceeding environmental parameter limits. After determining the fault category, fault characteristic parameters and the time of fault occurrence are recorded to form fault information.

[0029] Then, using a pre-defined model, the propagation path of fault information is inferred to determine the root cause of the fault. The root cause of the fault refers to the starting operation node in the causal graph of the operation chain that triggers the fault propagation path. The causal graph of the operation chain is a directed graph used to logically associate the operation records of a smart device over a period of time with fault phenomena. The pre-defined model can be a causal relationship model, a Bayesian network, a deep learning / network model, etc. For example, a fault probability back-inference method based on a Bayesian inference network can locate the maximum likelihood fault source as the root cause by calculating the posterior fault probability of each node. Another example is using a Long Short-Term Memory network model to capture time dependencies, or using an attention mechanism to identify the input feature that contributes most to fault prediction, i.e., the root cause of the fault. Optionally, a causal relationship model is used to infer the propagation path of fault information to determine the root cause of the fault. Specifically, the causal relationship model is a causal graph constructed based on the historical operation records and historical state operation data of the smart device. The causal graph includes nodes and directed edges; nodes are operation nodes and fault nodes. Directed edges point from one node to another, with one node being the cause and the other the effect, forming a causal relationship. In this way, after a fault occurs, the corresponding fault node is identified in the causal relationship model based on the fault information. A reverse search is then used to find the fault propagation path, thereby locating the root cause of the fault. For example, the fault information indicates that the fault node is an abnormal air conditioning cooling condition. A reverse search determines the fault propagation path as: user increases temperature setpoint → compressor frequency changes → fan speed changes → abnormal cooling. Therefore, the root cause is determined to be the user-issued control command to increase the temperature setpoint.

[0030] After identifying the root cause of the fault, the corresponding control commands are rolled back and the commands issued by the smart device are adjusted. This rollback, also known as operation rollback, involves using software commands to undo the operation that caused the fault and its subsequent dependent operations after the root cause is located, restoring the smart device to its safe state before the fault occurred.

[0031] In detail, reversing the corresponding control commands involves canceling the control command existing in the control center and clearing the parameter configuration corresponding to the control command, thereby eliminating the cause of the fault at the system level. Simultaneously, adjusting the commands issued by the intelligent devices ensures that the intelligent devices return to a safe state, preventing the fault from spreading. For example, if the filament of an intelligent lamp blows, the adjusted commands include issuing a power-off isolation command to disconnect it from the circuit and prevent overload current from affecting other bulbs; it may also include sending a signal optimization command to the control center, switching channels, increasing signal strength, etc., to ensure the normal transmission of subsequent commands. In this way, by eliminating the cause of the fault and adjusting the state of the intelligent devices, fault isolation and system self-healing are achieved. Furthermore, it improves fault handling efficiency and ensures the normal operation and stability of the equipment's collaborative work.

[0032] The collaborative network system control method for intelligent devices provided in this disclosure collects the operating status data of the intelligent devices, preprocesses it, and then performs fault analysis. A causal relationship model is used to perform fault causal reasoning, accurately locating the fault propagation path and the root cause. Then, by reversing control commands and adjusting commands issued by the intelligent devices, the fault causes are eliminated and the state of the intelligent devices is adjusted. In this way, fault isolation and system self-healing are achieved, fault handling efficiency is improved, and the normal operation and collaborative working stability of the equipment are ensured.

[0033] Optionally, the preset model in step S103 is a causal relationship model, which is constructed in the following way: S201, the processor obtains the historical operation sequence information of the smart device. The historical operation sequence information includes the occurrence time, type, and parameter change data after the operation is executed for each operation.

[0034] S202, the processor calculates the causal strength coefficient of two adjacent operations in the historical operation sequence information.

[0035] S203, the processor treats each operation as a node, the causal strength coefficient is the weight of the edge corresponding to the adjacent operation, and the direction of the edge is from the earlier operation to the later operation, in order to construct a causal relationship model.

[0036] Here, the historical operation sequence information includes the operation sequence V={ o 1 ，o 2 ，……，on}, each operation o i Time of occurrence t oi Operation type and data on changes in device parameters after operation execution. For any two adjacent operations... o i and o j Calculate the time difference Δ between the two operations t ij The correlation between the results of the two operations is analyzed to determine their causal strength coefficient. A higher correlation between the results of the two operations indicates a stronger causal strength coefficient. C ij The stronger the effect, the more precise the determination. Specifically, a maximum time interval is set for determining whether adjacent operations have a potential causal relationship, i.e., a time window threshold T. This threshold is preset based on the operational response characteristics of the smart device. Within the time difference Δ between two operations... t ij If the time window threshold T is less than the time window threshold T, then the two operations are temporally related; otherwise, they are not temporally related.

[0037] When temporal correlation exists, the logical relationship between two adjacent operations is determined based on their operation types. Specifically, a logical relationship rule base for operation types can be constructed. The strength of the logical relationship between adjacent operations is determined through forward matching and reverse validation of rationality, thus determining whether a valid logical relationship exists. For example, parameter adjustment operations and device response operations are logically related, state switching operations and load control operations are logically related, while software upgrade operations and hardware control operations are not logically related.

[0038] Calculate the correlation degree between the results of two adjacent operations if they are logically related. r ij Through analysis of operations o i and o j Once the consistency of parameter changes is determined, a multi-dimensional analysis method can be employed, considering the parameter type. Dimensions can include one or more of trend, magnitude, and time response. Specifically, trend consistency analysis involves fitting the parameter change curve after the operation is performed to the continuous parameter, calculating the slope similarity k. sim For example, for a smart refrigerator, operation o i This results in a brightness slope of 20% / s, and the operation... o j This results in a current slope of 0.15 A / s. If both trends are upward, k sim=0.92. Amplitude consistency analysis is performed on step-type parameters to calculate the proportion 'a' of the parameter's change amplitude. sim For example, for a smart refrigerator, operation o i This caused the wind speed to change from level 2 to level 4, a 100% change; operation o j If the air volume changes from 5 m³ / h to 10 m³ / h, with a change of 100%, then a sim =1.0. Time response consistency analysis calculates the time delay similarity t from the execution of an operation to a significant change in parameters. sim For example, operations o i The response latency is 100ms, and the operation... o j The response latency is 120ms, and the preset latency threshold is 200ms, then t sim =1 - |100 - 120| / 200 = 0.9. The final result's correlation is... r ij Calculated using a weighted average: r ij =α×k sim +β×a sim +γ×t sim Where α+β+γ=1. The correlation coefficient is... r ij The higher the causal strength coefficient, the stronger the causal strength coefficient. C ij The stronger.

[0039] Thus, the node set V based on operations is { o 1 ，o 2 ，……，o n}, the set of causal edges E={( o 1→ o 2, C 12 ), ( o 2→ o 3, C 23 ), ..., ( o i → o j , C ij Construct a weighted directed graph model. Causality strength coefficient. C ij The weights of the edges corresponding to adjacent operations are given, and the direction of the edges is from the earlier operation to the later operation.

[0040] Optionally, step S201 involves calculating the causal strength coefficient of two adjacent operations in the historical operation sequence information, including: Calculate the time difference and result correlation between two adjacent operations; the result correlation is determined based on the consistency between the equipment parameter change data corresponding to the earlier operation and the equipment parameter change data corresponding to the later operation.

[0041] When the time difference is less than the time difference threshold and the correlation of the result is greater than the correlation threshold, the ratio of the correlation of the result to the time difference is used as the causal strength coefficient.

[0042] Here, computational operations o i and o j Time difference Δ t ij =| t oj - t oi | and calculate the correlation based on the preceding content. r ij Then, the causality strength coefficient is calculated based on the following formula. C ij : (1) Where T is the preset time window threshold, and R is the result correlation threshold. C ij When >0, determine the operation. o i and o j There is a causal relationship, and C ij The larger the value, the stronger the causal relationship; the higher the correlation between outcomes and the shorter the time interval, the stronger the causal strength. Therefore, the edge set E satisfies... C ij A directed edge with weight > 0 has an edge weight of . C ij The order of operations, such as earlier and later, is determined by the timing of the operations. It can also represent the causal relationship between two operations, with the earlier operation being the cause and the later operation being the effect.

[0043] Optionally, the preset model is a causal relationship model. In step S103, the processor uses the preset model to infer the fault information propagation path to determine the root cause of the fault, including: S131, the processor determines the corresponding node of the fault in the causal relationship model based on the fault information, and uses that node as the starting point for the search.

[0044] S132, the processor uses a weighted depth-first search algorithm to perform reverse tracing in the causal relationship model.

[0045] S133, when the processor traces back to a point where there is no preceding node, it obtains the fault information propagation path and determines that the cause of the fault is the operation of the initial node in the fault information propagation path.

[0046] Here, a reverse search is performed based on the causal relationship model constructed earlier. Starting from the node corresponding to the fault, the search traces back along the reverse link of the causal relationship model to determine the operation of the initial node in the fault propagation, i.e., the root cause of the fault. Fault information includes fault type, fault characteristic parameters, and fault occurrence time. Based on this fault information, nodes directly related to the fault are matched in the causal relationship model, and these nodes are used as the search starting point, extending the search direction backward from the fault to the initial operation. During the reverse tracing process, a weighted depth-first search (WDFS) algorithm is used to prioritize edges with high causal strength coefficients for backward extension, ensuring that the search process prioritizes strong causal paths and filters out weak correlation interference, improving search efficiency and accuracy. Specifically, the weighted depth-first search algorithm, when traversing the causal graph, not only considers the path depth but also introduces the causal strength coefficient as a weight, prioritizing the exploration of paths with stronger causal relationships.

[0047] When tracing back to the initial operation in the causal relationship model where there is no valid predecessor node, the final result is obtained through path organization and root cause extraction. Specifically, fault propagation path extraction involves organizing the edges in the fault propagation path set `path` in the order of initial operation → intermediate operation → fault node to form a complete fault propagation path. Fault root cause determination involves taking the operation corresponding to the earliest initial node (the node with no preceding operation node) in the fault propagation path as the fault root cause. This initial operation is the starting point of the entire causal chain, and its abnormal execution directly or indirectly leads to abnormal subsequent operations, ultimately causing equipment failure. For example, organizing the edges in `path` yields the fault propagation path as: Operation o1: User increases temperature setting → Operation o2: Compressor frequency adjustment → Operation o3: Indoor unit fan speed adjustment → Fault v f The problem is a cooling malfunction. The initial node is operation o1, therefore, the root cause of the fault is determined to be the user's initial operation of raising the temperature setting.

[0048] Optionally, in step S132, the processor performs reverse tracing in the causal relationship model using a weighted depth-first search algorithm, including: During the reverse tracing process, the processor prioritizes searching edges with a causal strength coefficient higher than a weight threshold. The causal strength coefficient is the weight of the edge between adjacent nodes in the causal relationship model; nodes and edges constitute the causal relationship model.

[0049] Here, a causal strength threshold, i.e., a weight threshold C, is set. thThis is used to filter valid causal relationships between edges. When the weight C of an edge... ij >C th When C is found to be an edge, the causal relationship corresponding to that edge is deemed valid and included in the search scope. ij ≤C th If the edge is weakly related, it is considered invalid and skipped directly. This avoids irrelevant operations interfering with the search path, improving search efficiency and accuracy.

[0050] A complete description of the reverse tracing process is provided, including: 1) Data structure initialization: Stack S is used to store nodes to be traversed, following the Last-In-First-Out (LIFO) principle to ensure a depth-first search logic. Initially, the search starting point, i.e., the faulty node v, is... f Push it onto the stack. The visited node set `visited` is used to record the nodes that have been visited, avoiding path loops or search redundancy caused by repeated traversals. In the initial state, v... f Added to the visited list. The fault propagation path set `path` stores the valid causal edges identified during the search process. Each edge records the predecessor node, successor node, and weight information, ultimately forming a complete fault propagation path. The predecessor node is the node that precedes the previous node, and the successor node is the node that follows the previous node.

[0051] 2) Weighted reverse search execution: When stack S is not empty, pop the top node v; where, initially pop v f Traverse node v f All predecessor nodes o; perform validity checks on each predecessor node o. If o is not in visited and the weight C of edge (o, v) is... uv >C th If the edge is found to be a valid causal link, then push the valid predecessor node o onto the stack S. Simultaneously, add the edge (o, v) to the fault propagation path set path, and add o to the visited set to avoid repeated traversal. Repeat the above steps until the stack S is empty, meaning there are no more valid predecessor nodes to traverse.

[0052] For example, based on the abnormal cooling fault of the air conditioner v f Let node C be an example. th =0.1. Pop up v f Then, traverse its predecessor node o3, and the edge (o3, v) f The weight of o3 is 0.293 > 0.1, so o3 is pushed onto the stack and added to visited. The path then adds o3 → v. f0.293. Pop o3, traverse its predecessor node o2, edge (o2, o3) weight 0.28>0.1, push o2 onto the stack and add it to the visited list, the path adds o2→o3, 0.28. Pop o2, traverse its predecessor node o1, edge (o1, o2) weight 0.16>0.1, push o1 onto the stack and add it to the visited list, the path adds o1→o2, 0.16. Pop o1, traverse its predecessor nodes, find no valid preceding node, stack S is empty, search terminates.

[0053] Furthermore, the implementation process of the embodiments of this disclosure is described in detail, taking into account various smart washing machine malfunction scenarios: In the constructed causal model of the smart washing machine, the operation nodes include operation o1: the user's APP sets the washing mode to heavy-duty wash, t1=09:00:00; operation o2: the washing machine control board starts the motor to run at high speed, t2=09:00:02; operation o3: the water level sensor detects the water level, t3=09:00:03; operation o4: the inlet valve adjusts the water inlet flow, t4=09:00:05; the corresponding edge weight is C. 12 =0.32、C 23 =0.27, C 34 =0.21; Preset causality strength threshold C th =0.2.

[0054] The smart washing machine is experiencing an overfilling water malfunction. Malfunction information: Actual water level exceeds the set water level by 30%. Malfunction occurred at time t. f =09:00:08. The fault characteristic parameters are directly related to the parameter changes in operation o4: adjusting the inlet flow rate of the inlet valve. Analysis of the fault information determines the fault type to be a parameter anomaly fault, with the characteristic parameters being water level exceeding the standard by 30% and occurrence time t. f =09:00:08.

[0055] The faulty node v f Match operation node o4 and calculate r f4 =0.9 (High consistency between influent flow rate adjustment and water level exceeding the standard), Δt f4 =3s≤T=10s, C f4 =0.9 / 3=0.3, establish o4→v f A directed edge with a weight of 0.3.

[0056] Determine with v f Starting from the search point, a weighted depth-first search is used to trace back in reverse.

[0057] Initialization: Stack S = [v f ], visited={v f}, path= .

[0058] The first time v popped up f : Traverse the predecessor node o4, edge (o4, v f Since the weight 0.3 > 0.2, push o4 onto the stack, visited = {v f ,o4},path={(o4→v f ,0.3)}; The second pop of o4: Traverse the predecessor node o3, the edge (o3, o4) has a weight of 0.21 > 0.2, push o3 onto the stack, visited={v f ,o4,o3},path={(o4→v f ,0.3),(o3→o4,0.21)}; The third pop of o3: Traverse the predecessor node o2, the edge (o2, o3) has a weight of 0.27 > 0.2, push o2 onto the stack, visited={v f , o4, o3, o2}, path={(o4→v f ,0.3),(o3→o4,0.21),(o2→o3,0.27)}; The fourth pop of o2: Traverse the predecessor node o1, the edge (o1, o2) has a weight of 0.32 > 0.2, push o1 onto the stack, visited={v f , o4, o3, o2, o1}, path={(o4→v f , o3), (o3→o4, 0.21), (o2→o3, 0.27), (o1→o2, 0.32)}; The fifth pop of o1: Traverse the predecessor node, find no valid preceding node, stack S is empty, and the search terminates.

[0059] The fault propagation path and root cause were determined, and the path was summarized as follows: o1: Setting the powerful wash mode → o2: High-speed motor operation → o3: Water level sensor detection → o4: Inlet valve adjustment → v f Excessive water intake malfunction. The initial node in the malfunction propagation path is o1, and the root cause is the user's app setting to "intensive wash" mode. Investigation revealed that the water flow threshold setting for the "intensive wash" mode was abnormal, causing the inlet valve to continuously supply a large flow of water, thus triggering the malfunction.

[0060] Optionally, in step S104, the processor reverses the corresponding control instructions based on the fault state and controls the intelligent device to issue adjustment instructions, including: The processor cancels the erroneous operation instructions related to the root cause of the fault, clears the erroneous parameter configuration, and cuts off the fault propagation chain. And / or, The processor dynamically optimizes device operating parameters to compensate for performance deviations caused by faults. And / or, When the original control path fails, the processor automatically switches to a backup communication path or control node to ensure control continuity. And / or, When backup devices, sensors, or modules are available, the processor activates them to replace the faulty components and achieve functional redundancy backup.

[0061] After the processor is repaired, the device's operating status and fault characteristic parameters are continuously monitored. If the fault characteristics disappear and the device operates stably for a preset time, the self-healing is considered successful. If the fault still exists, the self-healing is considered unsuccessful, and the quantum degradation protocol is triggered.

[0062] Here, after locating the root cause and propagation path of the fault through fault point identification technology, self-healing strategies such as operation reverse rollback are executed to attempt to eliminate the fault and restore all equipment functions. Specifically, based on the fault location results, all operations directly related to the root cause in the fault propagation path are extracted to form a reverse rollback operation list. All operations include the initial fault operation and subsequent related operations. Operations in the list are rolled back one by one according to the principle of "last operation, first cancellation". For command-type operations, a command cancellation signal is sent to the equipment controller to terminate the execution of incomplete operations. For parameter configuration-type operations, erroneous parameter configurations are cleared, restoring the default parameters or historically optimal parameters before the operation execution. During the cancellation process, the equipment status is monitored in real time. If any abnormalities such as parameter mutations occur, the rollback is immediately paused, and the system switches to parameter adaptive adjustment mode. After the erroneous operation command rollback is completed, it is checked whether the key nodes in the fault propagation path have returned to normal, ensuring that the impact of the operation corresponding to the root cause of the fault has been completely eliminated. For example, a smart air conditioner experiences a compressor overload fault due to a user's misoperation in setting an excessively high temperature threshold. The fault propagation path is: incorrect temperature threshold setting → compressor high-frequency operation → overload alarm. When performing a reverse rollback, first cancel the compressor's high-frequency operation command, then clear the erroneous temperature threshold, and restore the default temperature setting to cut off the triggering conditions for the overload fault from the source.

[0063] Dynamically optimizing equipment operating parameters is used to address faults where there are no operational errors but parameter configurations are unreasonable, or performance deviations are caused by equipment aging or environmental changes. By dynamically optimizing operating parameters, functional degradation caused by the fault is compensated. Based on fault location results, core operating parameters related to the fault are determined, such as compressor frequency and fan speed for refrigeration faults, and damper opening and fan speed for ventilation faults. An appropriate optimization algorithm is selected based on the parameter type and fault characteristics. For linear parameters, a PID adaptive algorithm is used to dynamically adjust the proportional coefficient, integral time, and derivative time to ensure the parameter output closely matches the equipment's optimal operating range. For nonlinear parameters, a reinforcement learning algorithm is used to learn the optimal parameter adjustment strategy through real-time interaction with the equipment's operating environment. For multi-parameter collaborative optimization scenarios, a multi-objective optimization algorithm is employed to balance functional performance with constraints such as energy consumption and equipment load.

[0064] Control path reconfiguration is applicable to faults where the original control path fails, such as communication link interruption or control node failure. By switching to a backup path or node, the device control function is ensured to remain uninterrupted. Based on the fault location results, the type of failure in the original control path, such as communication link interruption, control node failure, or signal interference, and the location of the failure are determined.

[0065] When a communication link fails, the system automatically switches to a preset backup communication link, such as switching from Wi-Fi to Bluetooth, and updates the control signal transmission routing table to ensure normal transmission of control commands. When a control node fails, a backup control node is activated, such as switching from a master controller to a slave controller, and the control status and parameter configuration of the master node are synchronized using a consensus algorithm. A "connect first, switch later" strategy is adopted during the switchover process to avoid abrupt changes in device state caused by control interruption. After the switchover is complete, the system monitors indicators such as communication latency, signal strength, and packet loss rate of the backup path. If the indicators meet the control requirements, the current path is maintained; if instability still exists, redundant resources are activated.

[0066] Redundancy resource activation is used to address functional interruptions caused by hardware failures or the failure of a single component. By enabling backup resources, the function of the faulty component can be replaced. For example, if a smart washing machine's water level sensor malfunctions and cannot detect the water level, redundancy resource activation enables a backup water level sensor, synchronizes historical water level calibration parameters, and replaces the faulty sensor to perform the water level detection function, ensuring the washing process proceeds normally.

[0067] In this way, based on the fault location results, specific repair methods are matched for different fault types, avoiding the problem of low efficiency caused by indiscriminate repair.

[0068] Combination Figure 4 As shown, this disclosure provides a method for constructing a collaborative network system for smart devices, including: S101, the processor collects the operating status data of the smart device and preprocesses the collected operating status data.

[0069] S102, the processor performs fault analysis on the preprocessed running status data, and if a fault is found, it determines the fault information.

[0070] S103, the processor uses a preset model to infer the path of fault information propagation to determine the root cause of the fault.

[0071] S104: Based on the root cause of the fault, the processor reverses and rolls back the corresponding control instructions, and controls the intelligent device to issue adjustment instructions.

[0072] S105, when analyzing the situation of continuous faults, the processor obtains the function set, safe operation condition set, fault state set and condition judgment function of the intelligent device; the fault state set includes all existing fault states, and each function in the function set is matched with its corresponding priority.

[0073] S106, the processor sorts all functions in the function set according to their priority; the higher the priority, the earlier the corresponding function is sorted.

[0074] S107, the processor extracts the set of safe operating conditions corresponding to each function, calls the condition judgment function to compare the conflict between each function and the corresponding set of safe operating conditions, and eliminates the functions with conflict in the function set.

[0075] S108, the processor optimizes the set of functions that only contain non-conflicting cases, and controls the smart device to operate in a downgraded manner after each function in the optimized set meets the preset quantum security level requirements.

[0076] Here, when a smart device experiences a series of failures, a quantum degradation protocol is triggered to retain core functions while discarding fault-related functions. This ensures the system operates stably and safely in the degraded state, preventing secondary risks caused by the propagation of the fault. The function set F represents all the functions {f1, f2, ..., f...} possessed by the smart device. m This set is the object of the downgraded filtering. Function priority Pi, the order in which each function is ranked according to its importance, determines the order of checks during filtering. Safe operating condition set Cf. i It is function f i Necessary conditions for stable operation {c i1 c i2 , ..., c in For example, the washing function requires the water level sensor to be functioning correctly. The fault state set Fs represents all faults currently existing in the smart device {s1, s2, ..., s...}. kThis set is the core basis for verifying security conditions. The verification logic of the judgment function isSatisfied() is that when Cf i If none of the conditions in the condition conflict with Fs, return true and the function can be retained.

[0077] The steps to trigger the quantum degradation protocol include: according to functional priority P i The function set F is sorted from highest to lowest, prioritizing the verification rights of core functions. The sorted functions are traversed, and the `isSatisfied()` function is used to determine if the function's safe operating conditions conflict with the current fault state. Functions that pass verification are added to the retainable set `retainedF` until the set meets the system's quantum security level requirements. The intelligent device ultimately only runs functions in `retainedF`, discarding fault-related functions that cannot meet the security conditions. The quantum security level requirements are the minimum security standards for degraded system operation, including data transmission encryption strength, function access control permissions, and fault isolation boundaries. The conflict determination rule is: if any fault `s` exists in the fault state set Fs... k This will cause function f i The set of safe operating conditions Cf i At least one condition c in If the conditions cannot be met, it is considered a conflict; otherwise, it is considered a non-conflict. Furthermore, the set of safe operating conditions for the function, Cf... i It is necessary to specify the threshold values ​​for specific equipment components or parameters. For example, if the motor is normal, it is necessary to specify that the motor operating current is within the range of 0.5A to 2A and the speed fluctuation range is ≤5% to avoid vague descriptions that may lead to verification failure.

[0078] Taking a smart washing machine as an example, the fault point determination technology uses causal reasoning to locate the fault state Fs = {s1 (water level sensor fault), s2 (heating wire fault)}, triggering a quantum degradation protocol. Based on the Fs verification function, the washing function f1 fails to meet safety conditions due to the water level sensor fault, and the drying function f2 fails to meet safety conditions due to the heating wire fault. Ultimately, the smart reservation function f3 and the child lock function f4 are retained, achieving stable degraded operation. Here, continuous faults in smart devices refer to faults exhibiting a chain reaction, persistence, or concurrency at the logic control or system state level. In some embodiments, when continuous faults occur, a reverse rollback can be attempted. If the continuous faults cannot heal themselves, the quantum degradation protocol is triggered.

[0079] Optionally, in step S108, the processor optimizes the function set containing only non-conflicting cases by: The processor verifies the compatibility between functions in the set of functions that only contain non-conflicting cases and removes redundant functions to obtain an optimized set of functions.

[0080] Here, optimization is performed on the set of functions that do not conflict, ensuring the compatibility and security of the functions within the set. Compatibility verification refers to checking whether there are operational conflicts between functions in retainedF. For example, if some auxiliary functions depend on core functions that have been discarded, they must be discarded simultaneously. Redundant function elimination means that if multiple functions in retainedF achieve the same goal, the functions with higher priority or lower resource consumption are retained to reduce the operating load on the equipment.

[0081] Optionally, after step S102, the method further includes: The processor uses a directed acyclic graph to store the evolution history of the UI layout, and combines Δ encoding difference comparison to achieve chained version storage compression.

[0082] When the processor experiences a version compatibility failure or the smart device ceases maintenance, it will automatically downgrade to the corresponding management version according to preset rules.

[0083] Here, the evolution history of the DAG (Directed Acyclic Graph) storage UI layout is adopted, and efficient storage compression is achieved by comparing Δ encoding differences, including: Construct a DAG topology with the device's default UI version as the root node. Iterative version nodes point to a unique parent node, while repair version nodes point to multiple parent nodes, forming a version evolution DAG graph without circular dependencies. Each node contains a version number, version type, release timestamp, UI layout configuration data, dependent version node ID, and verification hash value.

[0084] Difference extraction and Δ encoding: Based on the parent version's UI layout configuration data, a structured comparison algorithm is used to extract the differences between the child versions, including changes in control attributes, additions or deletions of controls, and changes in interaction logic. A custom binary format is used to Δ encode the differences. Multi-level compressed storage: The Δ-encoded difference data is compressed in real-time using the LZ4 algorithm, while historical version difference data that has not been accessed for a long time is deeply compressed using ZIP64. Simultaneously, a difference data cache pool is set up to cache frequently accessed version differences.

[0085] Automatic downgrading to a suitable version based on preset rules includes: 1) Quantitative determination of downgrading trigger conditions. Downgrading timing is determined through multi-dimensional trigger rules, including equipment problem triggering and maintenance stoppage triggering. A multi-index weighted judgment model is used, and downgrading is triggered when the comprehensive score is ≥ a preset threshold. 2) Optimal downgrading version selection. Candidate downgrading versions are selected from the DAG version history based on preset screening rules. The Analytic Hierarchy Process (AHP) is used to score and rank the candidate versions, and the version with the highest score is selected as the target downgrading version. After pre-verification through sandbox simulation, the final downgrading version is determined. 3) Downgrading execution and rollback: Backing up the current version configuration data and user personalized settings, restoring the target downgrading version based on the DAG structure, and adapting and migrating the backed-up personalized settings to the downgrading version. After downgrading, the device's operating status is continuously monitored, supporting secondary downgrading due to faults and rollback after patch release. During version switching, the switching interruption time is ensured to be <50ms and the configuration drift rate <0.1%, including preloading the target version's core resources, using a double-buffered architecture to perform atomic switching, and controlling configuration drift through configuration consistency checks.

[0086] Optionally, S101, after the processor collects the operating status data of the smart device, it also includes: The processor collects and analyzes device energy consumption data to build a device energy consumption profile library.

[0087] The processor uses an energy consumption profile database as its core data foundation, integrates information from multiple sources to generate and execute power-saving strategies.

[0088] Here, energy consumption data of intelligent devices is collected under different operating states, time periods, and environmental conditions. This data includes core energy consumption parameters such as real-time power, cumulative power consumption, operating voltage, and operating current, along with corresponding timestamps, device operating modes, environmental parameters, and user operation commands. The collected raw energy consumption data is cleaned and standardized to remove noise and generate a structured dataset. Based on this preprocessed structured energy consumption dataset, a device energy consumption profile library is constructed using a classification modeling approach, enabling precise extraction and structured storage of energy consumption characteristics. Specifically, the data is categorized by device type and functional modules to ensure the profiles are relevant. Each device energy consumption profile includes basic device information, typical operating condition energy consumption characteristics, energy consumption fluctuation patterns, and high-energy-consumption scenario identifiers. A profile library update cycle is set, incorporating preprocessed energy consumption data from the previous day and updating energy consumption fluctuation patterns and high-energy-consumption scenario identifiers to ensure the profile library reflects real-time changes in device energy consumption.

[0089] Integrating multi-dimensional information and clarifying energy consumption optimization objectives provides a basis for strategy generation. This multi-dimensional information includes grid peak-valley time period division information, historical equipment usage data, and current equipment operating status. Based on the fusion analysis results, differentiated energy consumption optimization objectives are determined. For grid peak periods, the priority objective is to reduce the operating load of high-energy-consuming equipment and balance grid pressure. For high-frequency user usage periods, the priority objective is to moderately reduce energy consumption while ensuring user experience. For industrial equipment clusters, the priority objective is to balance the load and reduce overall energy costs.

[0090] Based on the fusion analysis results and optimization objectives, corresponding power-saving strategies are generated. These strategies include: 1) Off-peak operation strategy: Combining high-energy-consumption scenario identifiers in the energy consumption profile database with grid peak and valley periods, the operating time of high-energy-consumption equipment is scheduled to the grid valley period; simultaneously, user habits are matched to ensure that off-peak scheduling does not affect normal user operation. 2) Power limiting strategy: For non-core functional equipment or low-load demand scenarios, the maximum operating power is dynamically limited. For example, smart lights are limited to less than 5W when there is sufficient light, and air conditioners are limited to 80% of their rated cooling power during non-high-temperature periods; the power limiting threshold is determined based on the equipment energy consumption characteristics in the energy consumption profile database to avoid excessive limitation leading to functional failure. 3) Load balancing strategy: For equipment cluster scenarios such as multiple household air conditioners, the operating load is evenly distributed based on the energy consumption characteristics of each device in the energy consumption profile database. For example, the total cooling demand is distributed across multiple air conditioners to avoid a single air conditioner operating at high load for extended periods; the speed of multiple fans is adjusted to minimize overall energy consumption while meeting ventilation requirements.

[0091] Taking a home intelligent lighting system as an example, the embodiments of this disclosure will be described, wherein the home intelligent lighting system consists of multiple intelligent bulbs and a control center.

[0092] Data Acquisition and Preprocessing: The smart bulbs and the control center's device status sensors collect data in real time, such as bulb brightness, color temperature, and on / off status, as well as the control center's signal strength and command execution records. The data cleaning and feature extraction submodule processes this data, removing abnormal fluctuations and extracting key features, such as the bulb brightness change trend and the success rate of control command execution. For example, during data cleaning, if a sudden, large jump in bulb brightness data is detected that does not conform to physical laws (e.g., from a normal 50% brightness to 1000%), it is identified as noise data and removed. In terms of feature extraction, the rate of change of bulb brightness over a period of time is calculated as a key feature by analyzing the changes in bulb brightness.

[0093] Fault Analysis: The fault detection submodule detects that the brightness of a smart bulb suddenly drops to zero, indicating a fault. The fault classification submodule, based on relevant data, determines that it is a hardware fault of the bulb, such as a blown filament. The fault detection submodule may use a threshold detection method, for example, setting the bulb brightness below a certain threshold for a certain period of time, then classifying it as a brightness abnormality fault. The fault classification submodule further analyzes the bulb's current and voltage data, as well as the feedback information from control commands. If it finds that the current is zero and the voltage is normal, based on the bulb's operating principle, it determines that it is a hardware fault such as a blown filament.

[0094] Causal Reasoning: Previous operations were recorded, including user adjustments to bulb brightness via a mobile app and brightness adjustment commands sent by the control center. A causal graph was constructed based on these operations, revealing a potential correlation between the brightness adjustment operation and the current bulb malfunction. A reverse search of the causal graph using the WDFS algorithm determined that signal interference during the brightness adjustment command transmission caused a momentary overload in the bulb's internal circuitry, leading to filament burnout. Assuming the operation chain record shows that after the brightness adjustment command was issued, the bulb current momentarily increased beyond the normal range, and subsequently, the bulb brightness dropped to zero, a causal strength coefficient was calculated during causal graph construction based on the causal strength coefficient algorithm. Starting from the fault node where the bulb brightness was zero, the WDFS algorithm prioritized searching along edges with larger causal strength coefficients, revealing that signal interference during the brightness adjustment command transmission was the key factor causing the malfunction.

[0095] Operation Reverse Rollback: Based on causal reasoning, the module attempts to undo the brightness adjustment operation (although actual hardware failures cannot be fully recovered through software, this avoids other potential problems caused by the operation) and sends a command to the control center to re-establish a stable connection, ensuring that other bulbs are not affected. For example, the operation reverse rollback module sends a command to the control center requesting the retransmission of the previous brightness adjustment command, while simultaneously checking the circuitry connected to the affected bulb to ensure that no other potential faults have been transmitted to other bulbs.

[0096] Version Management: Stores the version history of the intelligent lighting system, including UI layout evolution and software function updates. When a bulb malfunction is detected that may be incompatible with the current version, the system version is downgraded to the previous stable version according to preset rules. The version switching control submodule quickly completes the version switch, ensuring that the switching interruption time is extremely short and does not affect the user's control of other bulbs. Assuming the current system version is V3, after detecting a bulb malfunction and a possible compatibility issue with the brightness adjustment algorithm of version V3, the automatic downgrade strategy submodule determines to downgrade the version to V2 according to preset version downgrade rules, such as prioritizing the most recent and thoroughly tested stable version. The version switching control submodule completes the version switch in a very short time through an optimized switching process. For example, it uses preloading technology to prepare key data for version V2 in advance, ensuring that the switching interruption time is less than 50ms. At the same time, through strict configuration management, it ensures that the configuration drift rate is less than 0.1%.

[0097] Energy Consumption Scheduling: Collects energy consumption data for each smart bulb, such as power consumption at different brightness levels. Based on the energy consumption profile database and users' daily usage habits, dynamically generates power-saving strategies, such as automatically reducing bulb brightness or disabling some bulbs during periods of sufficient daylight. The system sends energy consumption strategy commands to the smart bulbs, which execute the commands and return the results. The system determines the success of the operation based on the results; if successful, it continues to optimize the strategy; if unsuccessful, it re-analyzes the problem and adjusts the strategy. For example, the energy consumption profile database records that a smart bulb consumes 5W at 50% brightness and 10W at 100% brightness. The strategy generation submodule analyzes user habits and finds that users typically only need lower brightness lighting during the day. Therefore, when daylight intensity exceeds a certain threshold, it generates a power-saving strategy to reduce bulb brightness to 30%. The smart bulb receives the command, executes the operation, and returns the result. If the operation is successful, the system continues to monitor energy consumption data and further optimizes the strategy based on the actual situation, such as adjusting the brightness threshold according to seasonal changes. If the operation fails, the system re-analyzes the cause of the failure, such as whether the inability to adjust the brightness is due to command transmission errors or bulb hardware failure, and then adjusts the strategy or repairs the fault accordingly.

[0098] Combination Figure 5As shown, this disclosure provides a collaborative network system control device 100 for smart devices, including a processor 101 and a memory 102. Optionally, the device 100 may further include a communication interface 103 and a bus 104. The processor 101, communication interface 103, and memory 102 can communicate with each other via the bus 104. The communication interface 103 can be used for information transmission. The processor 101 can call logical instructions in the memory 102 to execute the collaborative network system control method for smart devices described in the above embodiment.

[0099] Furthermore, the logical instructions in the aforementioned memory 102 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0100] The memory 102, as a computer-readable storage medium, can be used to store software programs and computer-executable programs, such as program instructions / modules corresponding to the methods in the embodiments of this disclosure. The processor 101 executes functional applications and data processing by running the program instructions / modules stored in the memory 102, thereby implementing the collaborative network system control method for intelligent devices in the above embodiments.

[0101] The memory 102 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory 102 may include high-speed random access memory and may also include non-volatile memory.

[0102] Combination Figure 6 As shown, this disclosure provides a collaborative network system 200 for smart devices, including: The data acquisition and preprocessing module 201 is configured to acquire operating status data of the intelligent device and preprocess the acquired operating status data. The fault analysis module 202 is configured to perform fault analysis on the preprocessed operating status data, and determine fault information if a fault is found. The causal reasoning module 203 is configured to use a causal relationship model to reason about the propagation path of the fault information to determine the root cause of the fault. The reverse rollback module 204 is configured to perform reverse rollback on relevant control commands based on the root cause of the fault and control the intelligent device to issue adjustment commands.

[0103] The collaborative network system for smart devices also includes: a quantum degradation protocol module, configured to retain core functions and discard fault-related functions by triggering the quantum degradation protocol when the smart device cannot fully self-heal through reverse rollback. And / or, a version management module, configured to use a directed acyclic graph to store the UI layout evolution history, and combine Δ encoding difference comparison to achieve chained version storage compression; in the event of a version compatibility fault or the smart device ceasing maintenance, it automatically downgrades to the corresponding management version according to preset rules. And / or, an energy consumption management module, configured to collect and analyze device energy consumption data, build a device energy consumption profile library; using the energy consumption profile library as the core data foundation, it integrates multi-source information to generate and execute power-saving strategies.

[0104] This disclosure provides a computer-readable storage medium storing computer-executable instructions configured to execute the above-described collaborative network system control method for smart devices.

[0105] The technical solutions of this disclosure can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes one or more instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in this disclosure. The aforementioned storage medium can be a non-transitory storage medium, such as a USB flash drive, external hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, or other media capable of storing program code.

[0106] The foregoing description and accompanying drawings fully illustrate embodiments of this disclosure to enable those skilled in the art to practice them. Other embodiments may include structural, logical, electrical, procedural, and other changes. The embodiments represent only possible variations. Individual components and functions are optional unless explicitly required, and the order of operation may vary. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. Moreover, the terminology used in this application is for describing embodiments only and is not intended to limit the claims. As used in the description of embodiments and claims, the singular forms “a,” “an,” and “the” are intended to equally include the plural forms unless the context clearly indicates otherwise. Similarly, the term “and / or” as used in this application means including one or more of the associated listed items and all possible combinations thereof. Additionally, when used in this application, the term "comprise" and its variations "comprises" and / or "comprising" refer to the presence of stated features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof. Without further limitations, an element defined by the phrase "comprises a..." does not exclude the presence of other identical elements in the process, method, or apparatus that includes said element. In this document, each embodiment may focus on the differences from other embodiments, and similar or identical parts between embodiments can be referred to mutually. For methods, products, etc., disclosed in the embodiments, if they correspond to the method section disclosed in the embodiments, the relevant parts can be referred to the description of the method section.

[0107] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this disclosure. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0108] The methods and products disclosed in the embodiments herein (including but not limited to devices and equipment) can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units may be merely a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be electrical, mechanical, or other forms. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to implement this embodiment according to actual needs. In addition, the functional units in the embodiments of this disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0109] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

Claims

1. A control method for a collaborative network system of intelligent devices, characterized in that, include: Collect operational status data from smart devices and preprocess the collected operational status data; Fault analysis is performed on the preprocessed operating status data, and fault information is determined when faults are found. Using a pre-defined model, the propagation path of the fault information is inferred to determine the root cause of the fault. Based on the root cause of the fault, the corresponding control commands are reversed and rolled back, and the intelligent device is controlled to issue adjustment commands.

2. The method according to claim 1, characterized in that, The default model is a causal relationship model, which is constructed in the following ways: Obtain historical operation sequence information of the smart device, including the occurrence time, type, and parameter change data after each operation; Calculate the causal strength coefficient of two adjacent operations in the historical operation sequence information; Each operation is treated as a node, and the causal strength coefficient is the weight of the edge corresponding to the adjacent operation. The direction of the edge is from the earlier operation to the later operation, so as to construct a causal relationship model.

3. The method according to claim 2, characterized in that, Calculating the causal strength coefficient between two adjacent operations in the historical operation sequence information includes: Calculate the time difference and result correlation between two adjacent operations; the result correlation is determined based on the consistency of the device parameter change data corresponding to the earlier operation and the device parameter change data corresponding to the later operation. When the time difference is less than the time difference threshold and the correlation of the result is greater than the correlation threshold, the ratio of the correlation of the result to the time difference is used as the causal strength coefficient.

4. The method according to claim 1, characterized in that, The preset model is a causal relationship model. Using the preset model, the fault information propagation path is inferred to determine the root cause of the fault, including: Based on the fault information, determine the corresponding node of the fault in the causal relationship model, and use that node as the starting point for the search; Using a weighted depth-first search algorithm to perform reverse tracing in a causal relationship model; When tracing back to a point where no preceding node exists, the fault information propagation path is obtained and the root cause of the fault is determined to be the initial operation in the fault information propagation path.

5. The method according to claim 4, characterized in that, Using a weighted depth-first search algorithm to perform reverse tracing in a causal relationship model, including: During the reverse tracing process, edges with causal strength coefficients higher than the weight threshold are prioritized for searching; The causal strength coefficient is the weight of the edge between adjacent nodes in the causal relationship model, and the nodes and edges constitute the causal relationship model.

6. The method according to claim 1, characterized in that, Based on the root cause of the fault, the corresponding control commands are rolled back in reverse, including: Cancel erroneous operation commands related to the root cause of the fault, clear erroneous parameter configurations, and cut off the fault propagation chain; and / or, Dynamically optimize the operating parameters of smart devices to compensate for performance deviations caused by faults; and / or, When the original control path fails, it automatically switches to the backup communication path or control node to ensure control continuity; and / or, In the presence of backup equipment, backup sensors, or backup modules, activate the backup equipment, backup sensors, or backup modules to replace the faulty components and achieve functional redundancy backup.

7. The method according to any one of claims 1 to 6, characterized in that, Also includes: In the case of continuous faults, the set of functions, safe operating conditions, fault states and condition judgment functions of the intelligent device are obtained. The fault state set includes all currently existing fault states, and each function in the function set is matched with a corresponding priority. Sort all functions in the function set according to their priority; the higher the priority, the higher the corresponding function is ranked. Extract the set of safe operating conditions corresponding to each function, call the condition judgment function to compare the conflict between each function and the corresponding set of safe operating conditions, and eliminate functions with conflict in the function set; The system optimizes the set of functions that only contain non-conflicting cases, and after each function in the optimized set meets the preset quantum security level requirements, it controls the smart device to operate in a downgraded manner.

8. The method according to claim 7, characterized in that, Optimization of the feature set that only includes non-conflicting cases includes: The compatibility between functions in the set of functions that only contain non-conflicting cases is verified and redundant functions are removed to obtain an optimized set of functions.

9. A control device for a collaborative network system of intelligent devices, comprising a processor and a memory storing program instructions, characterized in that, The processor is configured to execute, when running the program instructions, the collaborative network system control method for intelligent devices as described in any one of claims 1 to 7.

10. A collaborative network system for intelligent devices, characterized in that, include: The data acquisition and preprocessing module is configured to collect the operating status data of the smart device and preprocess the collected operating status data. The fault analysis module is configured to perform fault analysis on the preprocessed operating status data, and determine the fault information if a fault is found. The causal reasoning module is configured to use a preset model to reason about the propagation path of the fault information to determine the root cause of the fault. The reverse rollback module is configured to reverse rollback the relevant control commands based on the root cause of the fault, and control the smart device to issue adjustment commands.

Images