Intelligent home recommendation method and system based on cross-modal fusion

Abstract

The invention relates to an intelligent home recommendation method and system based on cross-modal fusion. The method comprises the steps that user modal data, environment modal data and equipment behavior modal data are collected in real time; converting the environment modal data into quantum state amplitude features through a quantum encoder, and obtaining user behavior phase features through a time sequence feature extractor; performing cross-modal fusion on the quantum state amplitude features and the behavior phase features based on quantum entanglement operation to generate a dynamic fusion vector; inputting the dynamic fusion vector into a hybrid decision engine, outputting a recommendation instruction set by the hybrid decision engine, and verifying the physical feasibility of the recommendation instruction set based on an equipment state knowledge graph; and generating and executing a final recommendation instruction according to the verification result. According to the invention, intelligent matching of the job seeker and the recruitment demand is realized, and the recruitment efficiency and accuracy are improved.

Description

Technical Field

[0001] This application relates to the technical field of smart homes, and in particular to a smart home recommendation method and system based on cross-modal fusion. Background Technology

[0002] With the rapid development of IoT and AI technologies, modern smart home systems are facing multiple challenges, including processing massive amounts of heterogeneous data, real-time decision-making and response, and protecting privacy and security. Traditional smart home recommendation systems mainly rely on classic machine learning algorithms to process multi-source data such as user behavior, environmental sensing, and device status. When the system needs to integrate multimodal data such as voice, image, and sensor data to achieve personalized services, it generally suffers from low efficiency in cross-modal data fusion and semantic gaps. Static model architectures are difficult to adapt to the continuous evolution of the home environment and the drift of user needs, resulting in poor dynamic adaptability of the system and delays in response to sudden scenarios. Summary of the Invention

[0003] To address the aforementioned technical challenges, this application provides a home smart recommendation method and system based on cross-modal fusion. By employing quantum entanglement operations, it achieves deep collaboration between the environment, user, and device, thereby improving the system's speed in handling unexpected scenarios and its dynamic adaptability.

[0004] The above-mentioned objective of this application is achieved through the following technical solution:

[0005] A smart home recommendation method based on cross-modal fusion includes the following steps:

[0006] Real-time acquisition of user modal data, environmental modal data, and device behavior modal data;

[0007] The environmental modal data is converted into quantum state amplitude features by a quantum encoder, and the user behavior phase features are obtained by a time-series feature extractor.

[0008] Based on quantum entanglement operations, the amplitude characteristics and behavioral phase characteristics of the quantum state are fused across modes to generate a dynamic fusion vector;

[0009] The dynamic fusion vector is input into the hybrid decision engine, which outputs a recommended instruction set and verifies the physical feasibility of the recommended instruction set based on the device state knowledge graph.

[0010] The final recommended instructions are generated and executed based on the verification results.

[0011] By adopting the above technical solutions, quantum entanglement operations reduce the latency of environmental and behavioral data fusion, making it suitable for millisecond-level response scenarios in smart homes. Through three-dimensional verification of the device state knowledge graph, the interception rate of erroneous commands is improved. For example, when a gas leak (sudden environmental change) and a user leaving home (abnormal behavior) occur simultaneously in the kitchen, the quantum fusion module generates a dynamic fusion vector of "safety risk". The decision engine outputs the command "close the gas valve + start ventilation". After the knowledge graph verifies the circuit load safety, it is executed immediately, solving the problems of cross-modal data fragmentation and response latency in sudden scenarios.

[0012] In a preferred embodiment of this application, the step of converting the environmental modal data into quantum state amplitude features using a quantum encoder, and simultaneously obtaining user behavior phase features using a time-series feature extractor, specifically includes the following steps:

[0013] It receives continuous physical quantity data streams from environmental sensors, including temperature, light intensity, and equipment operating status, and discretizes the physical quantity values ​​into a probability distribution range of environmental quantum states that can be processed by quantum computing;

[0014] Capture user action event sequences, including voice command trigger times and device interaction intervals, construct a behavior timeline, and extract user behavior phase features.

[0015] By adopting the above technical solutions, the quantization of physical quantities has achieved precise discretization of physical quantities such as temperature (0-40℃→16 qubits) and light intensity (0-1000 lux→8 qubits). Compared with traditional ADC sampling error, the sampling error is reduced. Behavioral pattern recognition can distinguish between "active adjustment" and "accidental touch" behavior by constructing an "operation interval time axis", thus reducing the misjudgment rate. For example, during the rainy season, the system quantizes the environmental data of "humidity > 75% for 12 consecutive hours" into a quantum state probability distribution, and at the same time captures the temporal characteristics of the user's behavior of "frequently turning the dehumidifier on and off".

[0016] In a preferred embodiment of this application, the step of performing cross-modal fusion of the quantum state amplitude features and behavioral phase features based on quantum entanglement operations to generate a dynamic fusion vector specifically includes the following steps:

[0017] The quantum state amplitude feature is loaded onto the first quantum feature carrier, and the behavioral phase feature is loaded onto the second quantum feature carrier. A quantum entanglement initialization operation is performed on the two carriers to establish a quantum state correlation channel.

[0018] Monitor real-time changes in behavioral phase characteristics, trigger phase rotation operations, and dynamically adjust entanglement strength based on environmental amplitude characteristics;

[0019] Using the quantum state amplitude characteristics of the first quantum characteristic carrier as the control bit, a conditional rotation is performed on the behavioral phase characteristics of the second quantum characteristic carrier to generate a fused quantum state;

[0020] By projecting the fused quantum state into the classical data space, the dual-mode carrying capability of the vector is verified, and a dynamic fused vector carrying the amplitude characteristics and behavioral phase characteristics of the quantum state is generated.

[0021] By adopting the above technical solutions, quantum entanglement improves the correlation coefficient between ambient temperature and user behavior, thereby enhancing the relevance of recommended instructions. The entanglement initialization operation establishes a cross-modal physical correlation channel, avoiding manual weight setting. Conditional rotation is performed on the behavioral phase characteristics of the second quantum feature carrier, improving the dynamic adaptability of behavioral characteristics to environmental states and realizing context-aware fusion. The hierarchical adjustment of entanglement strength enables flexible switching of fusion strategies when environmental changes occur.

[0022] In a preferred embodiment of this application, the step of monitoring real-time changes in the phase characteristics of the behavior, triggering a phase rotation operation, and dynamically adjusting the entanglement strength based on environmental amplitude characteristics specifically includes the following steps:

[0023] Continuously receive user behavior time-series feature streams and calculate the phase offset between adjacent time windows;

[0024] When a positive mutation pattern is identified, the phase accelerator is activated to perform a positive rotation, thereby increasing the priority of the feature vector in the decision-making process.

[0025] When a negative drift pattern is detected, the phase compensator is triggered to perform a reverse rotation, reducing the confidence weight of the feature vector.

[0026] The environmental amplitude characteristics are analyzed in real time, and the entanglement energy level is adjusted through a quantum coupler. When the amplitude fluctuation is lower than a preset threshold, the low entanglement strength is maintained, and the independence of each mode characteristic is preserved.

[0027] When a continuous amplitude step is detected, switch to high entanglement strength and force the dual-mode features to be updated synchronously.

[0028] By adopting the above technical solutions, the calculation of phase offset quantifies the degree of behavioral pattern change, avoiding subjective threshold setting. The forward / reverse rotation mechanism dynamically normalizes behavioral characteristics, eliminating time scale differences. Amplitude step detection can accurately capture the moment of environmental change and improve the sensitivity of sudden response.

[0029] In a preferred embodiment of this application, the step of using the quantum state amplitude characteristics of the first quantum feature carrier as a control bit to perform a conditional rotation on the behavioral phase characteristics of the second quantum feature carrier to generate a fused quantum state specifically includes the following steps:

[0030] Extract the environmental data stability index and burst event marker intensity information from the quantum state amplitude features to generate a quantum control signal bound to the quantum state amplitude features;

[0031] The quantum control signal is input into the second quantum feature carrier interface. The rotation mechanism is activated according to the control signal type. The steady-state control signal triggers micro-amplitude phase calibration, and the sudden change suppression signal activates large-angle phase reset.

[0032] Perform rotation operations to enhance the weight of behavioral features through forward rotation and suppress abnormal behavior interference through reverse rotation;

[0033] During the rotation operation, quantum entanglement is maintained synchronously to generate a fused quantum state containing a main feature layer, a modulation layer, and a correlation layer. The main feature layer inherits the ambient amplitude distribution, the modulation layer carries the phase information after rotation, and the correlation layer records the interaction trajectory of the two carriers.

[0034] By adopting the above technical solutions, the stability indicators of environmental data and the intensity information of sudden event markers are extracted, the decision value of environmental characteristics is quantified, invalid signal interference can be avoided, the rotation mechanism is activated according to the control signal type, the intensity of behavior adjustment is dynamically matched, accurate scene adaptation is achieved, and new features are generated while retaining the original information by fusing quantum states, thus breaking through the limitations of simple weighted fusion.

[0035] In a preferred embodiment of this application, the step of inputting the dynamic fusion vector into a hybrid decision engine, the hybrid decision engine outputting a recommended instruction set, and verifying the physical feasibility of the recommended instruction set based on a device state knowledge graph specifically includes the following steps:

[0036] Extract the environmental state code containing abnormal event markers and the user intent probability distribution containing the confidence of each behavior pattern from the dynamic fusion vector to generate a dual-channel input data stream;

[0037] The environmental state code is input into the context-aware module to identify the current home scene type and predict the environmental evolution trend.

[0038] Input the user intent probability distribution into the intent reasoning module to analyze the user's explicit operation needs, mine the user's implicit behavioral preferences, and output a set of recommended instructions with confidence.

[0039] The recommended instruction set is loaded into the symbol rule validator, and the device status knowledge graph is linked to perform three-dimensional verification to generate instruction feasibility tags. The three-dimensional verification includes electrical compatibility, spatial accessibility and safety compliance verification.

[0040] By adopting the above technical solutions, the physical and semantic layers of cross-modal features are decoupled by separating environmental state coding and user intent probability distribution. The context awareness module constructs a spatiotemporal matrix containing 12-dimensional environmental parameters (temperature, humidity, light, etc.) to predict environmental trends for the next 30 minutes. The intent reasoning module innovatively introduces knowledge graph embedding technology and successfully mines users' implicit needs by constructing an intent reasoning tree containing three layers of implicit semantics. Through the linkage between the symbol rule verifier and the knowledge graph, the verification coverage of the physical feasibility of the final instruction is significantly increased.

[0041] In a preferred embodiment of this application: the instruction feasibility markers include green instructions that pass all constraints, yellow instructions that violate soft constraints, and red instructions that violate hard constraints. The step of generating and executing the final recommended instruction based on the verification results specifically includes the following steps:

[0042] Based on the feasibility flag of the instruction, hierarchical processing is performed: green instructions directly enter the execution queue, yellow instructions trigger the constraint optimizer, and red instructions start the safety coverage engine.

[0043] Output the final recommended instructions with correction records and execute them.

[0044] By adopting the above technical solution, a dynamic spectrum from safety to optimization is constructed using red, yellow and green markings, realizing visualized hierarchical management of risks. Red instructions correspond to hard constraints (such as electrical compatibility ≥95% pass rate), yellow instructions correspond to soft constraints (such as energy consumption exceeding the standard ≤15%), and green instructions correspond to full compliance status. This quantitative classification improves the transparency of system decision-making.

[0045] The second objective of this invention is achieved through the following technical solution:

[0046] A smart home recommendation system based on cross-modal fusion includes:

[0047] The data collection module is used to collect user modal data, environmental modal data, and device behavior modal data in real time.

[0048] The data processing module is used to convert the environmental modal data into quantum state amplitude features through a quantum encoder, and at the same time to obtain user behavior phase features through a time-series feature extractor.

[0049] A cross-modal fusion module is used to perform cross-modal fusion of the quantum state amplitude features and behavioral phase features based on quantum entanglement operations to generate a dynamic fusion vector;

[0050] The instruction recommendation module is used to input the dynamic fusion vector into the hybrid decision engine, the hybrid decision engine outputs a recommended instruction set, and verifies the physical feasibility of the recommended instruction set based on the device state knowledge graph;

[0051] The execution module is used to generate and execute the final recommended instructions based on the verification results.

[0052] The above-mentioned objective three of this application is achieved through the following technical solution:

[0053] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the aforementioned home smart recommendation method based on cross-modal fusion.

[0054] The fourth objective of this application is achieved through the following technical solution:

[0055] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned home smart recommendation method based on cross-modal fusion.

[0056] In summary, this application includes at least one of the following beneficial technical effects:

[0057] 1. This application transforms environmental data into quantum state amplitude features (e.g., a sudden temperature rise is mapped to a ground state probability of 0.95), encodes user behavior sequences into phase features (e.g., operation delays are marked as phase lag), and establishes a dynamic binding channel between the two carriers through quantum entanglement operations. Taking the scenario of an elderly person living alone falling as an example: when the gravity sensor detects 9.8 m / s 2 When an impact (sudden change in environmental amplitude) or a user's static timeout (behavioral phase lag) occurs, the system generates a strongly correlated fusion state, automatically triggering an alarm and floor heating, which improves the response speed and reduces the false alarm rate. The quantum amplitude-phase entanglement fusion mechanism solves the long-standing problem of the separation between environmental data and user behavior in the smart home field.

[0058] 2. This application dynamically adjusts the behavior phase through environmental amplitude characteristics: when a rainstorm is detected (continuous amplitude step), the sudden suppression signal triggers a 120° large-angle rotation, rotating the user's "open window" command phase from the ventilation demand zone to the safety protection zone, while simultaneously enhancing the entanglement strength to forcibly bind the "close window + dehumidify" command; in the energy optimization scenario, after the system identifies the peak electricity consumption (amplitude fluctuation marker), it performs a reverse rotation to reduce the weight of high-energy-consuming behavior commands. This quantum-level linkage mechanism between environment and behavior improves resource utilization efficiency and safety protection strength.

[0059] 3. By using a neural-symbolic hybrid decision engine and a three-dimensional knowledge graph verification system, the physical feasibility of recommended instructions is guaranteed. Traditional smart home systems often lead to dangerous operations due to neglecting real-world constraints, such as recommending "air conditioner to 16℃ strong cooling + oven start" which causes circuit overload. In this application, the initial instructions output by the decision engine need to be verified by the knowledge graph in three ways: electrical compatibility (load capacity verification), spatial accessibility (robotic arm path obstacle avoidance), and safety compliance (safety regulation review), which effectively reduces the equipment failure rate and reduces household electrical safety accidents. Attached Figure Description

[0060] Figure 1 This is a flowchart of an embodiment of a smart home recommendation method and system based on cross-modal fusion according to this application;

[0061] Figure 2 This is a flowchart of step S30 in an embodiment of a smart home recommendation method and system based on cross-modal fusion according to this application;

[0062] Figure 3 This is a flowchart of step S40 in an embodiment of a smart home recommendation method and system based on cross-modal fusion according to this application;

[0063] Figure 4 This is a schematic diagram of a smart home recommendation system based on cross-modal fusion, as described in this application.

[0064] Figure 5 This is a schematic block diagram of a computer device according to this application. Detailed Implementation

[0065] The present application will be further described in detail below with reference to the accompanying drawings.

[0066] In the example, such as Figure 1-3 As shown, this application discloses a smart home recommendation method based on cross-modal fusion, which specifically includes the following steps:

[0067] S10: Real-time acquisition of user modal data, environmental modal data, and device behavior modal data;

[0068] S20: The environmental modal data is converted into quantum state amplitude features by a quantum encoder, and the user behavior phase features are obtained by a time-series feature extractor.

[0069] S30: Based on quantum entanglement operations, cross-modal fusion is performed on the amplitude characteristics and behavioral phase characteristics of the quantum state to generate a dynamic fusion vector;

[0070] S40: Input the dynamic fusion vector into the hybrid decision engine, the hybrid decision engine outputs a recommended instruction set, and verifies the physical feasibility of the recommended instruction set based on the device state knowledge graph;

[0071] S50: Generate and execute the final recommended instructions based on the verification results.

[0072] In this embodiment, quantum entanglement reduces the latency of environmental and behavioral data fusion, making it suitable for millisecond-level response scenarios in smart homes. Through three-dimensional verification of the device state knowledge graph, the error command interception rate is improved. For example, when a gas leak (sudden environmental change) and a user leaving home (abnormal behavior) occur simultaneously in the kitchen, the quantum fusion module generates a dynamic fusion vector of "safety risk" within 0.2 seconds. The decision engine outputs the command "close the gas valve + start ventilation". After the knowledge graph verifies the circuit load safety, it is executed immediately, solving the problems of cross-modal data fragmentation and response latency in sudden scenarios.

[0073] In one embodiment, step S20 specifically includes the following steps:

[0074] S21: Receive continuous physical quantity data streams from environmental sensors, including temperature, light intensity and equipment operating status, and discretize the physical quantity values ​​into a probability distribution range of environmental quantum states that can be processed by quantum computing;

[0075] S22: Capture user operation event sequences, including voice command trigger times and device interaction intervals, construct a behavior timeline, and extract user behavior phase features.

[0076] Specifically, temperature values ​​are mapped to the probability of a qubit being in its ground state, light intensity is converted into the proportion of the |0> state in the quantum superposition state, and device on / off states are encoded as directional identifiers for quantum rotation operations. User operation event sequences are captured, including voice command trigger times and device interaction intervals, to construct a behavior timeline. Operation frequency distribution patterns are analyzed through a sliding window, key phase features of the behavior patterns are extracted, high-frequency operation clusters are identified as phase-leading features, operation delay events are converted into phase lag quantities, and periodic behavior patterns are manifested as periodic phase angle shifts.

[0077] In this embodiment, when the user turns on the reading light three times consecutively at 22:00 (behavioral phase characteristic: high-frequency stable operation), combined with the ambient brightness detected by the light sensor to be >90 lux (quantum amplitude characteristic: strong light probability 0.92).

[0078] In one embodiment, step S30 specifically includes the following steps:

[0079] S31: Load the quantum state amplitude feature into the first quantum feature carrier, load the behavioral phase feature into the second quantum feature carrier, perform a quantum entanglement initialization operation on the two carriers, and establish a quantum state correlation channel;

[0080] S32: Monitors real-time changes in behavioral phase characteristics, triggers phase rotation operations, and dynamically adjusts entanglement strength based on environmental amplitude characteristics;

[0081] S33: Using the quantum state amplitude characteristics of the first quantum characteristic carrier as the control bit, perform a conditional rotation on the behavioral phase characteristics of the second quantum characteristic carrier to generate a fused quantum state;

[0082] S34: Project the fused quantum state onto the classical data space to verify the dual-mode carrying capability of the vector and generate a dynamic fused vector carrying the amplitude characteristics and behavioral phase characteristics of the quantum state.

[0083] Specifically, the quantum state amplitude characteristics are loaded onto a first quantum feature carrier, and the behavioral phase characteristics are loaded onto a second quantum feature carrier. A quantum entanglement initialization operation is performed on the two carriers to establish a quantum state correlation channel. Real-time changes in the behavioral phase characteristics are monitored, triggering a phase rotation operation. Specifically, when a high-frequency operation cluster is detected, the second carrier is driven to perform a forward phase rotation; when an operation delay event occurs, the second carrier is driven to perform inverse phase compensation. The entanglement strength is dynamically adjusted based on the environmental amplitude characteristics. Specifically, when the environmental data is stable, a weakly entangled state is maintained, preserving mode independence; when the environment changes abruptly, a strongly entangled state is initiated. The dual-mode features are updated synchronously. Using the quantum state amplitude feature of the first quantum feature carrier as the control bit, the behavioral phase feature of the second quantum feature carrier is conditionally rotated to generate a fused quantum state. Its characteristics include: the amplitude component inherits the statistical distribution of environmental data, the phase component reflects the temporal change law of behavior, and the entanglement correlation degree characterizes the degree of environment-behavior matching. The amplitude component is mapped to the numerical intensity of the feature vector, the phase component is converted into the directional attribute of the feature vector, the amplitude intensity is checked to see if the environmental mutation mark is retained, the directional attribute is confirmed to contain the behavioral periodic pattern, and a dynamic fused vector carrying the joint features of environment and behavior is output.

[0084] In this embodiment, when window vibration (environmental change: abnormally enhanced amplitude characteristics) and the user's mobile phone moving away from the residence (behavioral characteristics: continuous phase lag) are detected, a strong entangled state binding dual features is triggered. The controlled door uses the window vibration amplitude as the control bit, rotates the behavioral phase to the warning range, and outputs a vector carrying the "high amplitude + negative phase" feature. The decision layer interprets this as an intrusion risk, executes window closing, and pushes an alarm.

[0085] In one embodiment, step S32 specifically includes the following steps:

[0086] S321: Continuously receive user behavior time-series feature streams and calculate the phase offset between adjacent time windows;

[0087] S322: When a positive mutation pattern is identified, the phase accelerator is activated to perform a positive rotation, increasing the priority of the feature vector in the decision-making process;

[0088] S323: When a negative drift pattern is detected, the phase compensator is triggered to perform a reverse rotation, reducing the confidence weight of the feature vector;

[0089] S324: Analyze the ambient amplitude characteristics in real time, adjust the entanglement energy level through a quantum coupler, maintain low entanglement strength when the amplitude fluctuation is lower than a preset threshold, and preserve the independence of each mode feature;

[0090] S325: When a continuous amplitude step is detected, switch to high entanglement strength and force the dual-mode features to be updated synchronously.

[0091] Specifically, a real-time phase tracking channel is established to continuously receive the time-series feature stream of user behavior, calculate the phase offset between adjacent time windows, and classify phase change patterns as follows: positive abrupt change mode: identifying events with a sudden increase in operation frequency; negative drift mode: detecting operation delays or interruptions; periodic oscillation mode: capturing regular behavioral cycle features. A rotation mechanism is activated based on the change pattern. When a positive abrupt change mode is identified, a phase accelerator is activated to perform positive rotation, increasing the priority of feature vectors in decision-making; when a negative drift mode is identified, a phase compensator is triggered to perform reverse rotation, reducing the confidence weight of feature vectors. The rotation angle is positively correlated with the phase offset, and the maximum rotation amplitude is limited by physical hardware constraints. Environmental amplitude features are analyzed in real time, and the amplitude fluctuation variance is calculated. When the amplitude fluctuation is below a preset threshold, it is marked as a steady-state environment. Amplitude step change points are identified, and continuous amplitude steps are marked as transient environments. An environment-entanglement intensity mapping rule is established to maintain low entanglement intensity in a steady-state environment and preserve the independence of each modal feature; in a transient environment, it switches to high entanglement intensity and forces the dual-modal features to be updated synchronously. The entanglement energy level is adjusted by a quantum coupler, and the intensity level is positively correlated with the amplitude of the sudden change.

[0092] In this embodiment, when the user leaves home (negative drift mode), the bit compensator performs a reverse rotation → reduces the weight of the "user at home" feature, and at the same time detects a sudden drop in room temperature of 5°C (amplitude step), switches the entanglement strength to a high energy level → forcibly binds the temperature change and the user leaving the field feature, and outputs a fusion vector to trigger the "energy saving mode": turn off the air conditioner and start low temperature protection.

[0093] In one embodiment, step S33 specifically includes the following steps:

[0094] S331: Extract the environmental data stability index and sudden event marker intensity information from the quantum state amplitude features, and generate a quantum control signal bound to the quantum state amplitude features;

[0095] S332: Input the quantum control signal into the second quantum feature carrier interface, activate the rotation mechanism according to the control signal type, trigger micro-amplitude phase calibration with steady-state control signal, and activate large-angle phase reset with sudden change suppression signal;

[0096] S333: Perform a rotation operation to enhance the weight of behavioral features through forward rotation and suppress abnormal behavior interference through reverse rotation;

[0097] S334: During the rotation operation, quantum entanglement is maintained synchronously to generate a fused quantum state containing a main feature layer, a modulation layer, and a correlation layer. The main feature layer inherits the ambient amplitude distribution, the modulation layer carries the phase information after rotation, and the correlation layer records the interaction trajectory of the two carriers.

[0098] Specifically, maintaining quantum entanglement synchronously during rotation includes: transmitting amplitude feature changes via a quantum bus and updating phase rotation parameters in real time. After generating the fused quantum state, quantum state characteristic detection is performed: verifying whether the main feature layer retains environmental mutation markers and detecting whether the modulation layer reflects behavioral pattern changes; and generating vectors through a quantum-classical converter: the amplitude component is mapped to numerical features, the phase component is converted into a direction vector, and the association layer information is encoded into feature confidence.

[0099] In this embodiment, when heavy rain is detected (sudden suppression signal: amplitude characteristics fluctuate violently), the control signal triggers a large-angle phase reset, rotates the user's "timed light off" behavior phase to emergency mode, and generates a fused state containing the main feature layer: heavy rain red warning mark; modulation layer: light keep-on command; association layer: environment-behavior strong coupling identifier, and output vector triggers the whole house lighting system to extend its operation.

[0100] In one embodiment, step S40 specifically includes the following steps:

[0101] S41: Extract the environmental state code containing abnormal event markers and the user intent probability distribution containing the confidence of each behavior pattern from the dynamic fusion vector, and generate a dual-channel input data stream;

[0102] S42: Input the environmental state code into the context awareness module to identify the current home scene type and predict the environmental evolution trend;

[0103] S43: Input the user intent probability distribution into the intent reasoning module, analyze the user's explicit operation needs, mine the user's implicit behavioral preferences, and output a set of recommended instructions with confidence.

[0104] S44: Load the recommended instruction set into the symbol rule validator, link the device status knowledge graph to perform three-dimensional verification, and generate instruction feasibility tags. The three-dimensional verification includes electrical compatibility, spatial accessibility and safety compliance verification.

[0105] In this embodiment, abnormal events are marked as sudden anomalies in environmental data (such as sudden temperature changes or smoke alarms), and the behavior pattern confidence level is a probabilistic quantification of the user's operational intent (such as "movie viewing mode" with a probability of 85%). Dual-channel input refers to the parallel processing channels that separate environmental and user data. Home scene types include security, comfort, and energy saving, and environmental evolution trends are predicted by changes in physical quantities, such as the rate of temperature rise. An embedded AI chip (such as NVIDIA Jetson) runs a lightweight neural network. Input: "Anomaly Type": "PM2.5 Exceeds Standard", "Value": 150 μg / m³ 3 Output: Scene Type = Health Alarm, Trend = Continuous Deterioration (Slope +12% / min). Explicit user operation needs include direct user commands (e.g., "adjust to 26℃"), while implicit behavioral preferences are derived from historical user habits (e.g., automatically turning off lights at 22:00). Electrical compatibility refers to the device power consumption / circuit load matching degree; exceeding limits is highlighted in red. Spatial accessibility refers to the feasibility of physical paths (e.g., robotic arm movement); obstacles are highlighted in yellow. Safety compliance refers to the degree of violation of safety regulations; violations of prohibitions are highlighted in red.

[0106] In one embodiment, the instruction feasibility markers include green instructions that pass all constraints, yellow instructions that violate soft constraints, and red instructions that violate hard constraints. Step S50 specifically includes the following steps:

[0107] S51: Perform hierarchical processing based on the feasibility flag of the instruction. Green instructions directly enter the execution queue, yellow instructions trigger the constraint optimizer, and red instructions start the safety coverage engine.

[0108] S52: Output the final recommended instructions with correction records and execute them.

[0109] In this embodiment, the initial recommendation for the (smart morning wake-up scenario) is: Main instruction: brew coffee + toaster; Alternative instruction: broadcast news. Knowledge graph verification: Electrical dimension: total power of breakfast appliance group exceeds limit (yellow mark); Safety dimension: coffee machine water tank is low (red mark). Conflict resolution: Power optimization: delay toaster start by 10 minutes; Safe replacement: coffee machine → instant hot water dispenser. Final execution sequence: "Instruction": "boil water", "Source": "safe replacement"; "Instruction": "broadcast news", "Source": "original alternative"; "Instruction": "toast", "Source": "delayed execution".

[0110] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0111] In one embodiment, a home smart recommendation system based on cross-modal fusion is provided, which corresponds one-to-one with a home smart recommendation method based on cross-modal fusion in the above embodiments. For example... Figure 4 As shown, the system includes:

[0112] The data collection module is used to collect user modal data, environmental modal data, and device behavior modal data in real time.

[0113] The data processing module is used to convert the environmental modal data into quantum state amplitude features through a quantum encoder, and at the same time to obtain user behavior phase features through a time-series feature extractor.

[0114] A cross-modal fusion module is used to perform cross-modal fusion of the quantum state amplitude features and behavioral phase features based on quantum entanglement operations to generate a dynamic fusion vector;

[0115] The instruction recommendation module is used to input the dynamic fusion vector into the hybrid decision engine, the hybrid decision engine outputs a recommended instruction set, and verifies the physical feasibility of the recommended instruction set based on the device state knowledge graph;

[0116] The execution module is used to generate and execute the final recommended instructions based on the verification results.

[0117] For specific limitations regarding a home smart recommendation system based on cross-modal fusion, please refer to the limitations of a home smart recommendation method based on cross-modal fusion mentioned above, which will not be repeated here. Each module in the aforementioned home smart recommendation system based on cross-modal fusion can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0118] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a home intelligence recommendation method based on cross-modal fusion.

[0119] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a home smart recommendation method based on cross-modal fusion.

[0120] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements a home intelligence recommendation method based on cross-modal fusion.

[0121] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0122] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to several functional units or modules as needed, that is, the internal structure of the device can be divided into several functional units or modules to complete all or part of the functions described above.

[0123] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A smart home recommendation method based on cross-modal fusion, characterized in that, Including the following steps: Real-time acquisition of user modal data, environmental modal data, and device behavior modal data; The environmental modal data is converted into quantum state amplitude features by a quantum encoder, and the user behavior phase features are obtained by a time-series feature extractor. Based on quantum entanglement operations, the amplitude characteristics and behavioral phase characteristics of the quantum state are fused across modes to generate a dynamic fusion vector; The dynamic fusion vector is input into the hybrid decision engine, which outputs a recommended instruction set and verifies the physical feasibility of the recommended instruction set based on the device state knowledge graph. The final recommended instructions are generated and executed based on the verification results.

2. The home smart recommendation method based on cross-modal fusion according to claim 1, characterized in that, The step of converting the environmental modal data into quantum state amplitude features using a quantum encoder, and simultaneously obtaining user behavior phase features using a time-series feature extractor, specifically includes the following steps: It receives continuous physical quantity data streams from environmental sensors, including temperature, light intensity, and equipment operating status, and discretizes the physical quantity values ​​into a probability distribution range of environmental quantum states that can be processed by quantum computing; Capture user action event sequences, including voice command trigger times and device interaction intervals, construct a behavior timeline, and extract user behavior phase features.

3. The home smart recommendation method based on cross-modal fusion according to claim 1, characterized in that, The step of performing cross-modal fusion of the quantum state amplitude features and behavioral phase features based on quantum entanglement operations to generate a dynamic fusion vector specifically includes the following steps: The quantum state amplitude feature is loaded onto the first quantum feature carrier, and the behavioral phase feature is loaded onto the second quantum feature carrier. A quantum entanglement initialization operation is performed on the two carriers to establish a quantum state correlation channel. Monitor real-time changes in behavioral phase characteristics, trigger phase rotation operations, and dynamically adjust entanglement strength based on environmental amplitude characteristics; Using the quantum state amplitude characteristics of the first quantum characteristic carrier as the control bit, a conditional rotation is performed on the behavioral phase characteristics of the second quantum characteristic carrier to generate a fused quantum state; By projecting the fused quantum state into the classical data space, the dual-mode carrying capability of the vector is verified, and a dynamic fused vector carrying the amplitude characteristics and behavioral phase characteristics of the quantum state is generated.

4. The home smart recommendation method based on cross-modal fusion according to claim 3, characterized in that, The steps of monitoring real-time changes in the phase characteristics of the behavior, triggering phase rotation operations, and dynamically adjusting the entanglement strength based on environmental amplitude characteristics specifically include the following steps: Continuously receive user behavior time-series feature streams and calculate the phase offset between adjacent time windows; When a positive mutation pattern is identified, the phase accelerator is activated to perform a positive rotation, thereby increasing the priority of the feature vector in the decision-making process. When a negative drift pattern is detected, the phase compensator is triggered to perform a reverse rotation, reducing the confidence weight of the feature vector. The environmental amplitude characteristics are analyzed in real time, and the entanglement energy level is adjusted through a quantum coupler. When the amplitude fluctuation is lower than a preset threshold, the low entanglement strength is maintained, and the independence of each mode characteristic is preserved. When a continuous amplitude step is detected, switch to high entanglement strength and force the dual-mode features to be updated synchronously.

5. The home smart recommendation method based on cross-modal fusion according to claim 3, characterized in that, The step of using the quantum state amplitude characteristics of the first quantum feature carrier as a control bit to perform a conditional rotation on the behavioral phase characteristics of the second quantum feature carrier to generate a fused quantum state specifically includes the following steps: Extract the environmental data stability index and burst event marker intensity information from the quantum state amplitude features to generate a quantum control signal bound to the quantum state amplitude features; The quantum control signal is input into the second quantum feature carrier interface. The rotation mechanism is activated according to the control signal type. The steady-state control signal triggers micro-amplitude phase calibration, and the sudden change suppression signal activates large-angle phase reset. Perform rotation operations to enhance the weight of behavioral features through forward rotation and suppress abnormal behavior interference through reverse rotation; During the rotation operation, quantum entanglement is maintained synchronously to generate a fused quantum state containing a main feature layer, a modulation layer, and a correlation layer. The main feature layer inherits the ambient amplitude distribution, the modulation layer carries the phase information after rotation, and the correlation layer records the interaction trajectory of the two carriers.

6. The home smart recommendation method based on cross-modal fusion according to claim 1, characterized in that, The step of inputting the dynamic fusion vector into the hybrid decision engine, the hybrid decision engine outputting a recommended instruction set, and verifying the physical feasibility of the recommended instruction set based on the device state knowledge graph specifically includes the following steps: Extract the environmental state code containing abnormal event markers and the user intent probability distribution containing the confidence of each behavior pattern from the dynamic fusion vector to generate a dual-channel input data stream; The environmental state code is input into the context-aware module to identify the current home scene type and predict the environmental evolution trend. Input the user intent probability distribution into the intent reasoning module to analyze the user's explicit operation needs, mine the user's implicit behavioral preferences, and output a set of recommended instructions with confidence. The recommended instruction set is loaded into the symbol rule validator, and the device status knowledge graph is linked to perform three-dimensional verification to generate instruction feasibility tags. The three-dimensional verification includes electrical compatibility, spatial accessibility and safety compliance verification.

7. The home smart recommendation method based on cross-modal fusion according to claim 6, characterized in that, The instruction feasibility markers include green instructions that pass all constraints, yellow instructions that violate soft constraints, and red instructions that violate hard constraints. The step of generating and executing the final recommended instruction based on the verification results specifically includes the following steps: Based on the feasibility flag of the instruction, hierarchical processing is performed: green instructions directly enter the execution queue, yellow instructions trigger the constraint optimizer, and red instructions start the safety coverage engine. Output the final recommended instructions with correction records and execute them.

8. A smart home recommendation system based on cross-modal fusion, characterized in that, include: The data collection module is used to collect user modal data, environmental modal data, and device behavior modal data in real time. The data processing module is used to convert the environmental modal data into quantum state amplitude features through a quantum encoder, and at the same time to obtain user behavior phase features through a time-series feature extractor. A cross-modal fusion module is used to perform cross-modal fusion of the quantum state amplitude features and behavioral phase features based on quantum entanglement operations to generate a dynamic fusion vector; The instruction recommendation module is used to input the dynamic fusion vector into the hybrid decision engine, the hybrid decision engine outputs a recommended instruction set, and verifies the physical feasibility of the recommended instruction set based on the device state knowledge graph; The execution module is used to generate and execute the final recommended instructions based on the verification results.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the home smart recommendation method based on cross-modal fusion as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the home smart recommendation method based on cross-modal fusion as described in any one of claims 1 to 7.

Images