An intelligent home linkage system based on an internet of things
By adapting hardware components according to device categories and dynamically generating fingerprint templates, the problems of device function differences and insufficient environmental adaptation in smart home systems are solved, improving the accuracy and adaptability of linkage and reducing system costs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GANSU SHENGHONG CONSTR ENG CO LTD
- Filing Date
- 2025-08-20
- Publication Date
- 2026-04-17
AI Technical Summary
In existing smart home systems, significant differences in the functions of core devices and edge devices lead to insufficient accuracy in linkage triggering, redundant components result in cost waste, single feature recognition is easily affected by environmental interference and triggers falsely, and templates do not combine user behavior and environmental changes, resulting in linkage effects that are out of sync with needs.
Smart home devices are categorized into trigger-type core devices, response-type edge devices, and auxiliary environmental devices. Hardware components are adapted according to the category, and fingerprint templates are generated through a feature extraction module. Resonance intervals are constructed by combining environmental parameters, and the templates are dynamically updated to adapt to user behavior and environmental changes.
The hardware configuration that matches the device functions has been achieved, which has improved the accuracy and reliability of the linkage, reduced the system cost, and ensured the adaptability and intelligence level in different environments.
Smart Images

Figure CN120779769B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of Internet of Things (IoT) and smart home control technology, specifically to a smart home linkage system based on IoT. Background Technology
[0002] With the development of IoT technology, smart home interconnection systems have gradually become the core of home intelligence, enabling scenario-based services through coordinated responses between devices. However, existing technologies still have the following shortcomings in practical applications:
[0003] Firstly, the core equipment and edge equipment have significantly different functions, but are configured with the same components. This results in the core equipment's lack of features affecting the accuracy of linkage triggering, while the edge equipment suffers from cost waste due to redundant components.
[0004] Secondly, existing systems mostly rely on a single feature to identify device status, which is easily affected by environmental interference, leading to false triggering. Furthermore, they cannot distinguish between user-initiated operations and automatic device operation, and the linkage logic lacks robustness.
[0005] Third, the templates are mostly built based on fixed rules, without taking into account the dynamic changes in user behavior, and lack environmental parameter adaptation, which leads to a decrease in template matching degree under different temperature, humidity and signal environments, and the linkage effect is out of touch with actual needs.
[0006] Therefore, there is an urgent need for a smart home linkage system that can adapt hardware according to device type and dynamically adapt to the environment. Summary of the Invention
[0007] To address the shortcomings of existing technologies, this invention provides an Internet of Things-based smart home linkage system, which solves the problems of redundant configuration and insufficient environmental adaptability of existing smart home devices.
[0008] To achieve the above objectives, the present invention provides the following technical solution: a smart home linkage system based on the Internet of Things, comprising:
[0009] The feature extraction module categorizes smart home devices into trigger-type core devices, response-type edge devices, and auxiliary-type environmental devices, and adapts hardware components according to the category. For trigger-type core devices and response-type edge devices, features are extracted from three aspects: vibration, energy consumption, and signal interference, respectively, to obtain their corresponding labels.
[0010] The fingerprint construction module filters out valid trigger events caused by user-initiated operations from the tags of core trigger devices, and extracts the device trigger sequence, interval duration, and amplitude to obtain a combination of time-series elements. Based on the recurring pattern of the time-series element combination, it generates a fingerprint template for the corresponding scenario.
[0011] In the environmental resonance module, when a new valid triggering event occurs, the system starts with the triggering core device, retrieves templates containing the device ID, calculates the matching degree with the temporal elements of the template, and filters fingerprint templates with a matching degree greater than or equal to the matching degree threshold into the candidate fingerprint template set. Environmental parameters are extracted from the tags of the triggering core device, and combined with the generated fingerprint template, a correlation baseline of three elements—environmental parameters, template ID, and core device function—is established. For each fingerprint template, the resonance interval between the core environmental parameters and the secondary environmental parameters is defined, and the templates in the candidate fingerprint template set are sorted according to the resonance interval.
[0012] As a further aspect of the present invention, the specific operation of adapting hardware components by category is as follows:
[0013] For core devices that are triggered, a built-in miniature vibration sensor, energy consumption monitoring chip, and signal coupler are included.
[0014] For responsive edge devices, only one component required for interaction with the core device is built-in;
[0015] For auxiliary environmental devices, no physical feature acquisition components are configured; only environmental parameters are output as a linkage reference.
[0016] As a further aspect of the present invention, feature extraction from three aspects—vibration, energy consumption, and signal interference—includes: in terms of equipment vibration, extracting the fundamental frequency and amplitude; in terms of equipment energy consumption, extracting the power fluctuation amplitude and duration; and in terms of signal interference, extracting the interference intensity and range of influence.
[0017] As a further aspect of the present invention, the tag format for trigger-type core devices and response-type edge devices is as follows:
[0018] The tags for triggering core devices are {[Device ID], [Actual values of six characteristics], [Compliance status of six characteristics], [Environmental parameters]};
[0019] The label for a response-type edge device is {[Device ID], [Status of collectable features], [Associated trigger-type core device ID]}, where the associated trigger-type core device ID is used to identify its response object.
[0020] As a further aspect of the present invention, for features acquired in real time, a threshold range corresponding to each feature is preset. If a feature is not within the corresponding threshold range, it indicates that the feature does not meet the standard, represented by 0; otherwise, it indicates that the feature meets the standard, represented by 1.
[0021] As a further aspect of the present invention, when a trigger-type core device generates a tag, the system first checks the number of compliant features among the six features. If at least three of them are compliant, the linkage can be triggered; otherwise, the linkage will not be triggered. For response-type edge devices, if all the collectable features are compliant and the status is 1, the linkage will be triggered; otherwise, the linkage will not be triggered.
[0022] As a further aspect of the present invention, the specific steps for filtering out valid triggering events caused by user-initiated operations are as follows:
[0023] Only retain device-triggered events that can be directly controlled by the user, excluding devices that can run automatically;
[0024] The tags of trigger-type core devices must meet at least four of the six characteristics;
[0025] Assign a unique event ID to each valid trigger event, in the format {[Device ID], [Valid Trigger Sequence Number]}.
[0026] As a further aspect of the present invention, the specific operation for generating a fingerprint template corresponding to a given scenario is as follows:
[0027] If the same combination of time-series elements appears ≥3 times, a fingerprint template for the corresponding scenario will be automatically generated.
[0028] The specific structure of the fingerprint template is: scene + temporal element combination + confidence level. In the temporal element combination, the interval duration and amplitude sequence are changed to a specified threshold range, and the confidence level is the number of repetitions / total number of events.
[0029] The repeated combination of the same time-series element includes the same device sequence, and the interval duration and amplitude are both within the specified threshold range;
[0030] The dynamic update mechanism for fingerprint templates includes:
[0031] Calculate the matching degree between the newly emerging temporal element combination and the existing fingerprint template. If the matching degree is ≥80%, increase the template confidence.
[0032] If the matching degree is less than 50% and the repetition is greater than or equal to 2 times, a new template will be generated.
[0033] As a further aspect of the present invention, the specific steps for calculating the matching degree between newly emerging temporal elements and existing templates are as follows:
[0034] Calculate order similarity =1 - (edit distance / number of template devices);
[0035] Edit distance represents the minimum number of operations required to convert the device sequence of the new timing sequence into the template sequence, including insertion, deletion, and replacement.
[0036] Calculate the similarity of a single interval = max(0, 1 - |new interval duration - center of template interval duration threshold interval| / center of template interval duration threshold interval), to obtain the interval duration similarity. =The average of the similarities of all individual intervals;
[0037] According to the formula Calculate amplitude similarity ;
[0038] Calculate pairwise collaborative scores for device sequence, interval duration, and amplitude sequence similarity:
[0039] The order and duration combined score is ;
[0040] The score for sequence and amplitude coordination is ;
[0041] The duration and amplitude were combined to score the following: ;
[0042] According to the formula Calculate the overall similarity ,in, The value range of is [0,1].
[0043] As a further aspect of the present invention, the specific operation of sorting the candidate fingerprint template set according to the resonance interval is as follows:
[0044] Each fingerprint template is labeled with the range of environmental parameters at the time of its generation, and this range is used as the template environment gene.
[0045] Based on the core device function of the trigger class corresponding to the template, determine one core environment parameter, and treat the others as secondary environment parameters;
[0046] Based on the real-time environmental parameters of the trigger-type core device, the templates in the candidate fingerprint template set are sorted:
[0047] The real-time core environment parameters of the triggering core device must fall within the core environment parameter range of the fingerprint template; otherwise, it will be directly rejected.
[0048] According to the formula Calculate the deviations of core environmental parameters, secondary environmental parameters, and corresponding resonance intervals. , ;
[0049] According to the formula Calculate the total deviation of environmental parameters, where, , These are the weights of the core environmental parameters and the secondary environmental parameters, respectively. ;
[0050] Sort the fingerprint templates in ascending order of total deviation. The smaller the deviation, the better the fit. Select the fingerprint template with the highest total deviation to execute the linkage.
[0051] The resonance interval of the core environmental parameter is taken as the 90% confidence interval of the core environmental parameter in the template environmental gene, and the resonance interval of the secondary environmental parameter is taken as the 80% confidence interval of the secondary environmental parameter in the template environmental gene.
[0052] This invention provides a smart home linkage system based on the Internet of Things, which has the following advantages compared with the prior art:
[0053] (1) This invention adapts hardware components according to device function classification, configures multi-dimensional perception components for triggering core devices to ensure the accuracy of linkage initiation, and configures simplified components for response and auxiliary devices to avoid redundancy. This satisfies the core devices’ need for full feature support and reduces the overall system cost.
[0054] (2) Based on the multi-feature cross-validation mechanism, this invention sets a threshold for the physical features extracted from the core devices of the trigger class, and triggers linkage by meeting at least three features, effectively filtering out misjudgments caused by fluctuations in a single feature or environmental interference. At the same time, combined with user-active operation filtering rules, it significantly improves the reliability of linkage judgment.
[0055] (3) This invention dynamically generates and updates fingerprint templates and constructs resonance intervals in combination with environmental parameters, so that the templates can not only adapt to changes in user habits, but also adapt to real-time environmental conditions, ensuring the adaptability of linkage strategies in different scenarios and improving the intelligence level of smart home systems. Attached Figure Description
[0056] Figure 1 This is the system principle block diagram of the present invention. Detailed Implementation
[0057] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0058] like Figure 1 This invention provides an Internet of Things-based smart home linkage system, comprising:
[0059] The feature extraction module categorizes smart home devices into trigger-type core devices, response-type edge devices, and auxiliary-type environmental devices, and adapts hardware components according to these categories.
[0060] Trigger-type core devices, such as washing machines, air conditioners, and smart door locks, have built-in miniature vibration sensors, energy consumption monitoring chips, and signal couplers;
[0061] Responsive edge devices, such as table lamps and curtain motors, only have one component built-in that is required for linkage with the core device;
[0062] Auxiliary environmental devices, such as temperature and humidity sensors and audible and visual alarms, are not equipped with physical feature acquisition components; they only output environmental parameters as linkage references.
[0063] If all devices adopt a uniform configuration, it is easy to waste costs. Core devices need to actively trigger linkages, requiring support from all dimensions of physical features; edge devices only need to respond to linkages, and a single component can meet the requirements.
[0064] For trigger-type core devices and response-type edge devices, feature extraction needs to be performed from three aspects: vibration, energy consumption, and signal interference.
[0065] The vibration waveform acquired by the miniature vibration sensor is subjected to Fast Fourier Transform (FFT), and the frequency with the highest energy percentage is taken as the fundamental frequency F.
[0066] The fundamental frequency is the frequency at which energy is most concentrated in a vibration signal. It is determined by the physical structure of the core components of the equipment. Different equipment has different core component structures, resulting in significant differences in fundamental frequencies.
[0067] In the time-domain waveform, the peak value of the vibration acceleration is taken, and then calculated using the formula... Convert to amplitude A, where a is the peak acceleration;
[0068] Amplitude is the maximum value of vibration displacement, which directly reflects the intensity of vibration. When the user actively operates the equipment, its amplitude is strongly correlated with the operating force and the operating status of the equipment, while the amplitude of external interference is usually weaker.
[0069] Regarding equipment energy consumption, the amplitude and duration of power fluctuations were extracted:
[0070] The energy consumption monitoring chip can collect the real-time power of the device and record the average power in the first 3 seconds before startup as the initial power. The power surge value after startup is taken as the peak power. ;
[0071] According to the formula Calculate the power fluctuation amplitude;
[0072] The core characteristic of equipment going from standby / shutdown to operation is a sudden change in power. In contrast, power changes caused by circuit interference or slight changes in equipment status are usually more gradual. The amplitude of power fluctuation is quantified as a relative value for this sudden change, which directly reflects whether the equipment has entered a substantial operating state.
[0073] From power exceeding Start timing from the beginning and stop timing when the power stabilizes within ±5% of the peak value, and record the duration T.
[0074] Once the equipment enters a stable operating state, the power will remain relatively stable within a certain range for a relatively long period of time. The duration of this stability is obtained to quantify the time during which the power stabilizes within the range after fluctuations, thus eliminating the influence of accidental interference.
[0075] Regarding signal interference, the interference intensity and range of influence are extracted:
[0076] The signal coupler monitors the received power of Wi-Fi signals around the device in real time. Record the average signal strength before the device starts up. Minimum signal strength after startup According to the formula Calculate the interference intensity I;
[0077] Signal interference intensity is a direct reflection of its ability to disrupt the communication / sensing functions of a device. Different devices have different tolerance thresholds for interference, and intensity quantification can directly determine whether the interference exceeds the device's tolerance limit.
[0078] Several reference points were set up around the equipment, and the interference intensity at each point was measured. The maximum distance where the interference intensity was greater than or equal to the interference threshold was taken as the influence range R.
[0079] The propagation of signal interference is affected by factors such as spatial attenuation and obstruction by obstacles. Its range of influence determines which devices will be affected. Local interference only involves nearby devices, while global interference will affect the entire system.
[0080] The core device that triggers the linkage needs to record its own status completely through tags. Therefore, the tag format is {[Device ID],[Actual values of six characteristics],[Six characteristics meet the standard],[Environmental parameters]}.
[0081] For the six features obtained, a threshold range is preset for each feature. If a feature is not within the corresponding threshold range, it indicates that the feature does not meet the standard and is represented by 0; otherwise, it indicates that the feature meets the standard and is represented by 1.
[0082] When a core device generates a tag, the system will first check the number of compliant features among the six features. If at least three of them are compliant, the linkage can be triggered; otherwise, the linkage will not be triggered.
[0083] Changes in the actual state of equipment are often accompanied by coordinated changes in multiple characteristics, rather than isolated fluctuations of a single characteristic. By using the rule of "three or more characteristics meeting the standard together", which includes any two aspects of equipment vibration, energy consumption, and signal interference, the linkage judgment is more in line with the physical laws of equipment operation.
[0084] The environmental parameters in the tag reference the output of auxiliary environmental devices, such as temperature and humidity;
[0085] For example, for a washing machine (ID: W01), its label can be represented as:
[0086] {[W01],[150,0.5,800,6,12,3],[1,1,1,1,0,0],[Temperature 25℃, Humidity 40%]};
[0087] As responders in the linkage, edge devices do not need to record complete features, but only their own collectable features. The label format is {[Device ID],[Collectible Feature Compliance Status],[Associated Trigger Core Device ID]}, where the associated trigger core device ID is used to identify its response object.
[0088] For example, a desk lamp containing only an energy consumption monitoring chip (response type, ID: L01) is tagged as: {[L01],[1,1],[W01]}. It meets the two characteristics of energy consumption fluctuation amplitude and duration. The associated triggering core device is washing machine W01, that is, it responds to the linkage command of W01.
[0089] If all collectable features meet the criteria, the system will respond and link the system; otherwise, it will not respond and link the system.
[0090] Auxiliary environmental devices have no data acquisition components and only output environmental parameters, so they do not generate tags themselves. However, the tags of trigger-type core devices will actively reference their parameters as environmental conditions.
[0091] The fingerprint building module filters out valid trigger events caused by user-initiated operations from the tags of core trigger devices, excluding automatic device operation or environmental interference. Specific rules include:
[0092] Only device-triggered events that can be directly controlled by the user are retained, such as smart door locks, washing machines, air conditioners, and light switches, while automatically operating devices such as refrigerators and water heaters are excluded;
[0093] Exclude irrelevant devices and interference events to ensure that the timing fingerprint only reflects the user's active behavioral habits, such as the user manually turning on the air conditioner, rather than the air conditioner automatically adjusting the temperature.
[0094] The triggering event must meet at least four of the six criteria, which is stricter than the basic linkage's requirement of at least three, to ensure that the user takes the initiative and avoids including non-user behaviors such as accidentally unlocking the door in the fingerprint.
[0095] Assign a unique event ID to each valid trigger event, in the format {[Device ID], [Valid Trigger Sequence Number]}, such as {[W01], [1]} for a door lock, which represents the first valid trigger of the door lock;
[0096] For the selected valid triggering events, three core timing elements are extracted: device triggering sequence, interval duration, and physical characteristic change amplitude. The specific operation is as follows:
[0097] Record the sequence of device IDs according to the order in which the events occur, such as door lock W01 → light L01 → air conditioner A01. Do not distinguish between multiple triggers of the same device. For example, light L01 → light L01 is considered invalid because the user will not operate the same light twice in a row.
[0098] Calculate the time difference between adjacent triggered actions, take a relative value, and do not record the specific time.
[0099] Non-digital recording avoids the leakage of user privacy; for example, it does not record when a user returns home at 18:00, but only records the sequence of events and the relative intervals.
[0100] For each triggered action, calculate the relative rate of change of the core physical characteristics:
[0101] If the triggering action targets a vibrating device, such as a door lock or washing machine, then the calculated amplitude = (actual amplitude / standard amplitude) × 100%;
[0102] If the trigger action targets energy-consuming devices, such as lights or air conditioners, then the calculated amplitude = (actual power fluctuation amplitude / standard fluctuation amplitude) × 100%;
[0103] If the trigger action targets signal interference, such as that of a radio frequency remote control, then the calculated amplitude = (actual interference intensity / standard interference intensity) × 100%;
[0104] The magnitude of changes in physical characteristics reflects the intensity and habits of user operations. For example, if a user habitually starts the air conditioner on the highest setting, with a magnitude of 120%, the fingerprint will record this characteristic.
[0105] The valid trigger event timing element combination obtained is {[device sequence], [interval duration], [amplitude sequence]};
[0106] For example, when a user comes home, they first open the smart door lock (amplitude 80%), then turn on the lights (amplitude 30%) after 5 seconds, and then turn on the air conditioner (amplitude 60%) after 10 seconds. The extracted time sequence elements are {[W01→L01→A01][5s,10s],[80%,30%,60%]}.
[0107] Based on the recurring patterns of temporal elements, a scene fingerprint template is automatically generated. The specific operation is as follows:
[0108] If the same combination of time-series elements appears ≥3 times, a fingerprint template for the corresponding scenario will be automatically generated, such as the home scenario or the bedtime scenario.
[0109] The repeated combination of the same time-series element includes the same device sequence, and the interval duration and amplitude are both within the specified threshold range;
[0110] The specific structure of the fingerprint template is: scene + temporal element combination + confidence level. In the temporal element combination, the interval duration and amplitude sequence are changed to a specified threshold range, and the confidence level is the number of repetitions / total number of events.
[0111] For example, scenario: Homecoming scenario + device sequence: W01→L01→A01 + interval duration threshold range: [5-0.5, 5+0.5], [10-1, 10+1], [80%-12%, 80%+12%], [30%-4.5%, 30%+4.5%], [60%-9%, 60%+9%] + confidence level 3 / 5;
[0112] The dynamic update mechanism for fingerprint templates specifically includes:
[0113] Calculate the matching degree between newly emerging time series elements and existing templates. If the matching degree is ≥80%, increase the template confidence.
[0114] If the matching degree is less than 50% and the repetition is greater than or equal to 2 times, a new template is generated, such as the homecoming scene - variant 1;
[0115] The specific steps for calculating the matching degree between newly emerging time series elements and existing templates are as follows:
[0116] Order similarity is calculated using edit distance normalization. That is, order similarity = 1 - (edit distance / number of template devices);
[0117] Edit distance refers to the minimum number of operations required to convert the device sequence of the new timing into the template sequence. The operations specifically include insertion, deletion, and replacement, with each operation scoring 1 point.
[0118] The device triggering order is a core feature of user habits. For example, "turning on the lights before turning on the air conditioner" and "turning on the air conditioner before turning on the lights" are completely different habits. Editing distance can accurately measure the difference in order.
[0119] The similarity of time intervals is calculated using the relative error average. For each adjacent interval, the similarity is calculated separately, and then the average value is taken.
[0120] Where, the similarity of a single interval = max(0,1-|new interval duration-template interval duration threshold interval center| / template interval duration threshold interval center), and the total duration similarity = the average of all single interval similarities;
[0121] The interval duration is a continuous value. Fluctuations in the time of user operation are normal. The relative error can reflect whether the fluctuation is within a reasonable range.
[0122] Calculating amplitude sequence similarity using cosine similarity The amplitude sequences of the template and the new time series are treated as vectors, and the similarity is measured by the angle between the vectors. The specific formula is as follows: ;
[0123] Amplitude reflects the intensity ratio of user operations. For example, turning on the air conditioner is usually more intense than turning on the lights. Cosine similarity is good at measuring the directional consistency of vectors rather than the difference in absolute value, which is consistent with the natural fluctuation characteristics of the intensity of user operations.
[0124] Calculate pairwise collaborative scores for device sequence, interval duration, and amplitude sequence similarity:
[0125] The order and duration combined score is ;
[0126] The score for sequence and amplitude coordination is ;
[0127] The duration and amplitude were combined to score the following: ;
[0128] In stable user behavior habits, the similarity of the three dimensions will show a synergistic effect of rising and falling together. For example, the more similar the order, the more similar the duration and magnitude tend to be. This synergy can be measured by the product of the similarity of each pair of dimensions. The larger the product, the more synchronized the similarity of the two dimensions is, and the stronger the synergy.
[0129] According to the formula Calculate the overall similarity ,in, The value range is [0,1]. The higher the value, the stronger the synergy among the three dimensions and the higher the habit matching degree.
[0130] The overall similarity is determined by the geometric mean of the three cooperability scores, because the geometric mean can amplify the influence of high cooperability and suppress the interference of low cooperability.
[0131] The environmental resonance module, when a new valid triggering event occurs, starts from the core device of the triggering class, searches for templates containing the device ID. For example, if door lock W01 is triggered, it searches for all templates containing W01, calculates the matching degree between the valid triggering event and the temporal elements of the template, filters fingerprint templates with a matching degree ≥ the matching degree threshold, and puts them into the candidate fingerprint template set.
[0132] Environmental parameters are extracted from the tags of the trigger-type core devices. These parameters may include temperature (T), humidity (H), sound wave (S), etc. Combined with the fingerprint template generated in the second step, a baseline is established to link the environmental parameters, template ID, and core device functions.
[0133] Each fingerprint template is marked with the range of environmental parameters generated at the time of its generation. For example, the environment when the "homecoming scene template" is generated is mostly "T∈[23,27], H∈[35,45], S<40dB", which serves as the template environment gene.
[0134] Based on the core device function of the trigger class corresponding to the template, determine a core environmental parameter. For example, in the "Homecoming Scene", the air conditioner is the core device and the core environmental parameter is T; in the "Bedtime Scene", the humidifier is the core device and the core environmental parameter is H, and others are secondary environmental parameters.
[0135] For each fingerprint template, based on environmental parameters and user historical operation data, the resonance range of core environmental parameters and secondary environmental parameters is determined. The specific operation is as follows:
[0136] For the resonance interval of the core environmental parameter, the 90% confidence interval of the core environmental parameter in the template environmental gene is taken. For example, the T gene of the "homecoming scene - summer" template is 26-30℃, and the 90% confidence interval is 27-29℃.
[0137] For the resonance interval of minor environmental parameters, the 80% confidence interval of minor environmental parameters in the template environmental genes is taken;
[0138] Confidence intervals ensure that the intervals cover the most frequently occurring environmental ranges of the template, thereby improving the matching accuracy between the intervals and the template;
[0139] Based on the real-time environmental parameters of the trigger-type core device, the templates in the candidate fingerprint template set are sorted. The specific operation is as follows:
[0140] The real-time core environment parameters of the triggering core device must fall within the core environment parameter range of the fingerprint template; otherwise, it will be directly rejected.
[0141] The core environmental parameter range of the fingerprint template ensures basic compatibility between the template and the current environment, preventing core devices from operating in unsuitable environments;
[0142] According to the formula Calculate the deviations of core environmental parameters, secondary environmental parameters, and corresponding resonance intervals. , ;
[0143] According to the formula Calculate the total deviation of environmental parameters, where, , The weights for core and secondary environmental parameters are respectively assigned and need to be adjusted based on the actual situation. ;
[0144] Sort the fingerprints in ascending order of total deviation. The smaller the deviation, the better the fit. Select the fingerprint template ranked first to execute the linkage.
[0145] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.
[0146] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.
Claims
1. An Internet of Things-based smart home linkage system, characterized by, include: The feature extraction module categorizes smart home devices into trigger-type core devices, response-type edge devices, and auxiliary-type environmental devices, and adapts hardware components according to the category. For trigger-type core devices, features are extracted from three aspects: vibration, energy consumption, and signal interference. For response-type edge devices, only the features that can be collected by themselves are extracted to obtain their corresponding labels. The fingerprint construction module filters out valid trigger events caused by user-initiated operations from the tags of core trigger devices, assigns a unique event ID to each valid trigger event in the format {[Device ID], [Valid Trigger Sequence Number]}, and extracts the device trigger sequence, interval duration, and amplitude according to the tags and format to obtain a combination of time sequence elements. Based on the recurring pattern of the time sequence element combination, a fingerprint template for the corresponding scenario is generated. The environmental resonance module, when a new valid triggering event occurs, starts with the triggering core device, retrieves fingerprint templates containing the device ID, calculates the matching degree of the combination of temporal elements of the fingerprint template, and filters fingerprint templates with a matching degree ≥ the matching degree threshold into the candidate fingerprint template set. It extracts environmental parameters from the tags of the triggering core device, and, combined with the generated fingerprint template, establishes a baseline for the association of three elements: environmental parameters, fingerprint template ID, and triggering core device function. For each fingerprint template, it calibrates the resonance interval between the core environmental parameters and the secondary environmental parameters, and sorts the fingerprint templates in the candidate fingerprint template set according to the resonance interval between the core environmental parameters and the secondary environmental parameters. The calibration of the resonance interval between the core environmental parameters and the secondary environmental parameters includes: for the resonance interval of the core environmental parameters, taking the 90% confidence interval of the core environmental parameters in the fingerprint template environmental gene; For the resonance interval of minor environmental parameters, the 80% confidence interval of minor environmental parameters in the fingerprint template environmental gene is taken; the fingerprint template environmental gene represents the environmental parameter interval at the time of generation of each fingerprint template. 2.The smart home linkage system based on the Internet of Things according to claim 1, wherein, The specific steps for adapting hardware components by category are as follows: For core devices that are triggered, a built-in miniature vibration sensor, energy consumption monitoring chip, and signal coupler are included. For response-oriented edge devices, only one component required for linkage with trigger-oriented core devices is built-in; For auxiliary environmental devices, no physical feature acquisition components are configured; only environmental parameters are output as a linkage reference. 3.The smart home linkage system based on the Internet of Things according to claim 1, characterized in that, Feature extraction is performed from three aspects: vibration, energy consumption, and signal interference. In terms of equipment vibration, the fundamental frequency and amplitude are extracted; in terms of equipment energy consumption, the power fluctuation amplitude and duration are extracted; and in terms of signal interference, the interference intensity and range of influence are extracted. The equipment includes trigger-type core equipment and response-type edge equipment. 4.The smart home linkage system based on the Internet of Things according to claim 1, wherein, The tag format for trigger-type core devices and response-type edge devices is as follows: The tags for triggering core devices are {[Device ID], [Actual values of six characteristics], [Compliance status of six characteristics], [Environmental parameters]}; The label for a response-type edge device is {[Device ID], [Status of collectable features], [Associated trigger-type core device ID]}, where the associated trigger-type core device ID is used to identify its response object. 5.The smart home linkage system based on the Internet of Things according to claim 4, characterized in that, For features acquired in real time, a threshold range is preset for each feature. If a feature is not within the corresponding threshold range, it indicates that the feature does not meet the standard, represented by 0; otherwise, it indicates that the feature meets the standard, represented by 1. 6.The smart home linkage system based on the Internet of Things according to claim 4, characterized in that, When a trigger-type core device generates a tag, the system first checks the number of compliant features among the six features. If at least three of them are compliant, the linkage can be triggered; otherwise, the linkage will not be triggered. For response-type edge devices, if all the collectable features are compliant and the status is 1, the linkage will be triggered; otherwise, the linkage will not be triggered. 7.The smart home linkage system based on the Internet of Things according to claim 1, wherein, The specific steps for filtering out valid trigger events caused by user-initiated actions are as follows: Only retain device-triggered events that can be directly controlled by the user, excluding devices that can run automatically; The tags of trigger-type core devices must meet at least four of the six characteristics; Assign a unique event ID to each valid trigger event, in the format {[Device ID], [Valid Trigger Sequence Number]}. 8.The smart home linkage system based on the Internet of Things according to claim 1, wherein, The specific steps for generating a fingerprint template for a given scenario are as follows: If the same combination of time-series elements appears ≥3 times, a fingerprint template for the corresponding scenario will be automatically generated. The specific structure of the fingerprint template is: scene + temporal element combination + confidence level. In the temporal element combination, the interval duration and amplitude sequence are changed to a specified threshold range, and the confidence level is the number of repetitions / total number of events. The repeated combination of the same time-series element includes the device triggering sequence being consistent, and the interval duration and amplitude being within the specified threshold range; The dynamic update mechanism for fingerprint templates includes: Calculate the matching degree between the newly emerging temporal element combination and the existing fingerprint template. If the matching degree is ≥80%, increase the fingerprint template confidence. If the matching degree is less than 50% and the fingerprint is repeated ≥ 2 times, a new fingerprint template will be generated. 9.The smart home linkage system based on the Internet of Things according to claim 8, wherein, The specific steps for calculating the matching degree between newly emerging temporal element combinations and existing fingerprint templates are as follows: Computing device trigger order similarity = 1 - (edit distance / number of fingerprint template devices); The edit distance represents the minimum number of operations required to convert the device triggering order of the new timing sequence into the fingerprint template order, including insertion, deletion, and replacement. Calculate the single interval similarity = max(0, 1 - |new interval duration - center of fingerprint template interval duration threshold interval| / center of fingerprint template interval duration threshold interval), to obtain the interval duration similarity. =The average of the similarities of all individual intervals; According to the formula Calculate the amplitude similarity ; Calculate pairwise collaborative scores based on device triggering sequence, interval duration, and amplitude similarity: The device trigger sequence is scored in conjunction with the interval duration ; Device trigger order and amplitude synergy score is ; The interval duration and amplitude synergy score is ; According to the formula The comprehensive similarity is calculated Wherein, The value interval of is [0, 1]. 10.The smart home linkage system based on the Internet of Things according to claim 1, wherein, The specific operation for sorting the candidate fingerprint template set based on the resonance interval between core environmental parameters and secondary environmental parameters is as follows: Each fingerprint template is labeled with the range of environmental parameters at the time of its generation, and this range is used as the fingerprint template environmental gene. Based on the core device function of the trigger class corresponding to the fingerprint template, determine one core environment parameter, and treat the others as secondary environment parameters; Based on the real-time environmental parameters of the trigger-type core device, the fingerprint templates in the candidate fingerprint template set are sorted: The real-time core environment parameters of the triggering core device must fall within the core environment parameter range of the fingerprint template; otherwise, it will be directly rejected. According to the formula calculating the deviation of the core environmental parameters, the secondary environmental parameters, and the corresponding resonance interval ; The total deviation of the environmental parameters is calculated according to the formula wherein, are the weights of the core environmental parameters and the secondary environmental parameters, respectively, and ; Sort the fingerprint templates in ascending order based on the total deviation of environmental parameters. The smaller the deviation, the better the fit. Select the fingerprint template ranked first to execute the linkage.
Citation Information
Patent Citations
Smart home interaction control method based on equipment with screen
CN113359497A
Linkage method, platform and equipment of smart home system and storage medium
CN113777947A