Remote control method and device for intelligent multimedia interaction equipment, equipment and medium

Abstract

The invention provides a remote control method and device for intelligent multimedia interaction equipment, equipment and a medium, which are used for improving the real-time performance and the execution success rate of remote control and enabling complex multimedia interaction to be reliably operated on a heterogeneous network and heterogeneous equipment. The method comprises the following steps: carrying out hierarchical analysis processing on interaction data input by a user to generate perception command data, and carrying out execution planning and resource allocation on the perception command data according to equipment performance characteristics to generate an executable operation sequence, and performing transmission quality evaluation and path selection on the propagation path network according to the executable operation sequence to obtain a target transmission path, and performing protocol encapsulation on the executable operation sequence according to the target transmission path to obtain a remote control instruction data stream.

Description

Technical Field

[0001] This application relates to the field of remote control technology, and in particular to a remote control method, apparatus, device and medium for an intelligent multimedia interactive device. Background Technology

[0002] The remote control function of multimedia interactive devices lies in integrating dispersed terminal devices into a unified control and management closed loop, thereby enabling remote content distribution and display, real-time interactive response, centralized operation and maintenance and fault diagnosis, and operation optimization based on usage data. Through remote control, operators can complete content updates, parameter tuning, access control and security auditing without touching the devices; users can obtain a collaborative experience across devices (such as collaborative meetings, exhibition linkage, and remote teaching), while the system can collect operational and interactive telemetry data for continuous improvement and predictive maintenance, improving service availability and operational efficiency.

[0003] Existing remote control methods for multimedia interactive devices mostly employ typical cloud / client or remote desktop models, including REST / HTTPS-based control APIs, MQTT / AMQP-based message push, command issuance via secure tunnels through VPNs or dedicated channels, and manual maintenance using remote desktops / VNC, SSH, etc. In addition, it's common to send voice or natural language requests to the cloud for intent parsing before issuing simple control commands, and to use OTA updates and centralized monitoring platforms for batch configuration and log collection. Security and reliability are typically ensured through TLS, certificates, HMAC, or hardware security modules, but these existing implementations are mostly simply stacked functional modules.

[0004] Existing remote control methods for multimedia interactive devices lack end-to-end latency sensitivity and device capability awareness coordination. Most solutions treat parsing, decision-making, and transmission as independent stages or process them centrally in the cloud, failing to perform pre-deployment based on the terminal's real-time status and resource adaptation before distribution, and failing to dynamically select transmission paths and error correction strategies as needed. Therefore, in network jitter or heterogeneous device environments, high perceived latency, command failures, or repeated retransmissions are prone to occur, resulting in low real-time interactive experience and execution success rate. Summary of the Invention

[0005] This application provides a remote control method, apparatus, device, and medium for intelligent multimedia interactive devices, which improves the real-time performance and execution success rate of remote control, enabling complex multimedia interactions to operate reliably on heterogeneous networks and devices.

[0006] In a first aspect, this application provides a remote control method for an intelligent multimedia interactive device, applied to a central control system, the method comprising: The user-input interaction data is processed in layers to generate perception command data; Based on the preset device performance characteristics, the perception command data is executed and resources are allocated to generate an executable operation sequence; Based on the executable operation sequence, the transmission quality of the preset propagation path network is evaluated and the path is selected to obtain the target transmission path; The executable operation sequence is protocol-encapsulated according to the target transmission path to obtain a remote control command data stream.

[0007] In one possible implementation, the hierarchical parsing and processing of user-input interaction data to generate perceptual command data includes: Based on the preset interaction control type, the user input interaction data is classified and processed to obtain first-level interaction data and second-level interaction data; According to the preset command mapping strategy, the first-level interactive data is converted into the first-level command; The second-level interactive data is semantically parsed and intent is identified to generate the second-level command.

[0008] In one possible implementation, the step of performing execution planning and resource allocation on the perceived command data based on preset device performance characteristics to generate an executable operation sequence includes: The preset parameters of the preset equipment performance characteristics are identified and extracted to obtain the set of target parameters of the equipment, and the resource constraint strategy of the corresponding equipment is generated according to the set of target parameters of the equipment. The perception command data is subjected to sandbox simulation processing based on the resource constraint strategy to obtain a pre-execution sequence. Based on the device performance characteristics, resource allocation and timing planning are performed on each operation step in the pre-execution sequence to generate an executable operation sequence.

[0009] In one possible implementation, the step of performing transmission quality assessment and path selection on a preset propagation path network based on the executable operation sequence to obtain a target transmission path includes: Traverse each path in the preset propagation path network to extract the current state parameters corresponding to each path; Calculate the transmission quality score of the corresponding path based on the set of timing constraints in the executable operation sequence and the current state parameters; Based on the transmission quality score, the paths in the propagation path network are prioritized for transmission, and a transmission execution path sequence is generated. The target transmission path is selected from the propagation path network according to the preset number of paths and the transmission execution path sequence, and the corresponding transmission feature data is obtained from the preset transmission feature database according to the target transmission path.

[0010] In one possible implementation, the step of protocol encapsulating the executable operation sequence according to the target transmission path to obtain a remote control command data stream includes: According to the encapsulation strategy, a transmission order tag is added to each data in the executable operation sequence to obtain an encapsulated data sequence; According to the transmission protocol, a checksum and a security token are added to each piece of data in the encapsulated data sequence to generate an encrypted dataset; Based on the transmission order tags corresponding to each data in the encrypted dataset, the encrypted dataset is sorted and framed to generate a control command data stream. Based on the transmission quality score corresponding to the target transmission path, forward error correction coding is added to each data frame in the control command data stream to generate a remote control command data stream.

[0011] In one possible implementation, the method further includes: The received real-time execution status data from the intelligent multimedia interactive device is decompressed to obtain execution feedback data; The device performance characteristics are dynamically calibrated and updated based on the execution feedback data using a preset dynamic adjustment method.

[0012] Secondly, this application provides a remote control device for an intelligent multimedia interactive device, the device comprising: The interaction parsing module is used to perform layered parsing and processing of user-input interaction data to generate perception command data; The planning and allocation module is used to perform execution planning and resource allocation on the sensing command data according to preset device performance characteristics, and generate an executable operation sequence. The path selection module is used to perform transmission quality assessment and path selection on the preset propagation path network according to the executable operation sequence, so as to obtain the target transmission path. The protocol encapsulation module is used to encapsulate the executable operation sequence according to the target transmission path to obtain a remote control command data stream.

[0013] In summary, this application includes at least the following beneficial technical effects: 1. By performing hierarchical parsing of input and collaborative processing in the cloud, it achieves instant response to millisecond-level real-time commands, significantly reducing perceived latency and improving the interactive experience.

[0014] 2. By using pre-execution simulation based on device performance, device-aware resource allocation, and transmission quality assessment, the matching of transmission and execution is achieved, reducing command failure rate and improving end-to-end reliability and bandwidth utilization.

[0015] 3. By implementing hierarchical security strategies and encapsulation, the confidentiality, integrity, and resilience to packet loss of commands are improved, thereby enhancing system security and auditability. Attached Figure Description

[0016] Figure 1 This is a diagram illustrating the operating environment of a remote control method for an intelligent multimedia interactive device provided in an embodiment of this application. Figure 2 This is a flowchart illustrating a remote control method for an intelligent multimedia interactive device executed by a central control system according to an embodiment of this application. Figure 3 This is a flowchart illustrating a remote control method for an intelligent multimedia interactive device performed by an intelligent multimedia interactive device according to an embodiment of this application. Figure 4 This is a schematic diagram of the structure of a remote control device for an intelligent multimedia interactive device provided in an embodiment of this application; Figure 5 This is a schematic diagram of the structure of the computing device provided in the embodiments of this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. With the development of technology and the emergence of new scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0018] The terms "first," "second," etc., used in the specification, claims, and drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the description of embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not expressly listed or inherent to those processes, methods, products, or apparatuses.

[0019] Please refer to the above. Figures 1 to 3 As shown: The operating environment of the remote control method for intelligent multimedia interactive devices provided in this application includes a user input acquisition device, a central control system located in the cloud, and the intelligent multimedia interactive device. The user input acquisition device is the physical entry point for user interaction with the intelligent multimedia interactive device, responsible for capturing the user's original commands. The user input acquisition device is located in a place convenient for direct user operation, such as in the user's hand (e.g., a smartphone, remote control), on a conference room table (e.g., a conference panel), or in an intelligent space covered by a camera / microphone array (e.g., the entire living room). User input acquisition devices typically integrate: a sensor array for capturing multimodal user interaction data, such as a microphone array for high-fidelity voice capture and beamforming and suppressing ambient noise (e.g., a 4- or 6-microphone array based on Infineon XENSIV™ MEMS microphones (e.g., the IM69D130); a touchscreen for receiving user input touch and swipe commands; and a depth-sensing camera for real-time capture of gestures and body posture (e.g., the Intel RealSense Tracking Camera T265 or the Orbbec Astra series); and a main control unit integrating data preprocessing and data transmission functions, where data preprocessing ensures data transmission quality and efficiency (e.g., audio noise reduction (algorithm processing, not semantic understanding), image compression (e.g., H.264 encoding), and data packetization).

[0020] The central control system, located in a cloud data center, includes a gateway with communication capabilities, a CPU / GPU cluster integrating multiple artificial intelligence models and rule engines, a dynamic database that stores and updates device performance characteristics in real time, and a security and transmission scheduler. The central control system performs the following steps: Step S11: Perform layered parsing processing on the user-input interaction data to generate perception command data.

[0021] It should be understood that interactive data is a collection of multimedia signals generated by the user through a user input acquisition device, specifically encompassing modalities such as voice waveforms, touch coordinate sequences, gesture contour data, and text character streams. Voice waveforms are captured by a microphone array, manifesting as a continuous analog signal with sound pressure varying over time, which is converted into a discrete digital sequence after analog-to-digital conversion. Touch coordinate sequences originate from the driving circuitry of a capacitive touchscreen, recording the two-dimensional coordinates and movement trajectory of the fingertip's contact position. Gesture contour data is acquired through a depth-sensing camera, including the three-dimensional coordinates and motion vectors of skeletal joints. Text character streams originate from key events on a keyboard or virtual input method. The acquisition of interactive data relies on the sensor arrays built into the user input acquisition device; for example, Infineon MEMS microphones are used for audio acquisition, Orbbec Astra cameras are used for visual data capture, and the device's main control chip, such as the ESP32-S3, is responsible for data preprocessing and packetization, before transmitting it to the cloud via a wireless network. Interactive data carries the user's original operational intent and forms the signal basis for remote control.

[0022] Upon receiving interactive data, to differentiate between real-time requirements and computational complexity and thus optimize resource allocation, the central control system first categorizes the data according to preset interactive control types. These types are predefined based on historical operation patterns and device performance models. For example, operations requiring millisecond-level response, such as touch swipes and gesture zooming, are categorized as Level 1 interactive data, while operations allowing for hundreds of millisecond delays, such as natural language queries and complex configuration modifications, are categorized as Level 2 interactive data. The classification process is implemented through a rule engine and lightweight feature extraction. For instance, touch data with a coordinate change rate exceeding a threshold is automatically marked as Level 1, while voice data with a spectral entropy value below a threshold is tentatively classified as Level 2 for further verification. For example, when a user rapidly swipes their finger on the touchscreen of a smart TV remote, the resulting touch coordinate sequence is identified as Level 1 interactive data due to its high instantaneous change rate; the user's voice waveform stating "dim the conference room lights and play jazz" is classified as Level 2 interactive data because it contains multiple intentions and a complex semantic structure. This categorization prevents high-real-time operations from being blocked by complex parsing processes while ensuring that non-real-time operations receive sufficient semantic analysis.

[0023] For the first-level interactive data after classification, the system converts it into first-level commands using a pre-defined command mapping strategy. This strategy, centered on lookup tables and state machines, directly maps low-level data patterns into device-executable atomic operations. The command mapping strategy is pre-defined based on the device interface specification; for example, the Delta value of a touch coordinate sequence is mapped to the displacement parameter of a scroll command, and the pinch angle of a gesture contour is mapped to the scaling factor of a zoom command. The conversion process does not involve semantic analysis, relying only on geometric features and kinematic models. For example, the tangent angle of the touch trajectory is calculated using the directional derivative formula: if the trajectory point set is... Then the instantaneous direction angle Used to determine the swipe direction and map it to the direction indicator in navigation commands. First-level interaction data is directly mapped to minimize parsing latency and meet the stringent requirements for instant feedback in interactive scenarios. For example, during video playback, when a user swipes up or down on the touchscreen, the system does not need to understand the semantics of "adjust volume"; it only needs to map the vertical displacement to the numerical parameters of the volume adjustment command to achieve imperceptible latency control.

[0024] For the classified second-level interactive data, the system performs semantic parsing and intent recognition processing to generate second-level commands. This embodiment employs a hybrid architecture combining deep learning models and knowledge graphs for parsing the second-level interactive data. Semantic parsing first fuses features from multimodal data; for example, speech waveforms are converted to time-frequency features using Mel-frequency cepstral coefficients, gesture contour data has spatial relationship features extracted using graph convolutional networks, and text character streams are converted into semantic vectors using word embedding models. Intent recognition is built on a Transformer architecture, and its attention weight calculation can be expressed as Attention(Q,K,V)=Softmax. V, where the query matrix Q represents the feature representation of the current interaction data, and the key matrix K and value matrix V come from the historical intent pattern library. Multi-dimensional similarity calculation is used to match predefined intent categories. For example, after the user's voice command "Find the sales report for the third quarter of last year" is automatically converted to text by speech recognition, the semantic parsing model identifies "find" as the operation type, "sales report" as the target object, and "third quarter of last year" as the time parameter, ultimately generating a composite command (i.e., a second-level command) that includes both database query and visualization rendering operations. Semantic parsing and intent recognition resolve the ambiguity of complex intents. For example, in "Play Jay Chou's 'Seven Mile Fragrance'", "Jay Chou" needs to be mapped to artist metadata, and "Seven Mile Fragrance" needs to be mapped to the song title. Named entity recognition and slot filling ensure the accuracy of the operation.

[0025] Finally, the obtained first-level and second-level commands are encapsulated to obtain perceptual command data. This perceptual command data is a set of tuples, with each command containing an opcode, parameter list, and priority label. The perceptual command data provides standardized instruction input, and its hierarchical mechanism allows the central control system to allocate computing resources on demand—first-level commands directly enter the fast execution channel, while second-level commands trigger global optimization strategies. For example, in a smart home scenario, when a user gestures to issue the command "close the curtains," the first-level command immediately generates a motor control signal through a mapping strategy; while when a user makes a voice request to "schedule the robot vacuum cleaner to clean the living room next Wednesday at 10 AM," the second-level command, after semantic parsing, generates a complex operation sequence containing time triggers and area planning parameters. This hierarchical parsing balances system response speed and functional integrity, ensuring both the smoothness of basic interactions and support for the implementation of advanced intelligent services.

[0026] Step S12: Based on the preset device performance characteristics, perform execution planning and resource allocation on the sensing command data to generate an executable operation sequence.

[0027] It should be understood that device performance characteristics, as a priori knowledge base in the planning phase, are essentially digital abstractions of the hardware capabilities and software interfaces of the controlled intelligent multimedia interactive device. Device performance characteristics encompass a set of parameters such as processor architecture, core frequency, memory bandwidth, graphics processing unit rendering capabilities, hardware codec support list, storage input / output performance, and peripheral interface protocols. The acquisition of these characteristics relies on technical specifications provided by the device manufacturer and historical performance benchmark data from the cloud. For example, the SoC model of a smart TV determines its upper limit of video decoding capabilities, and the memory capacity of a conference tablet limits the number of multimedia tasks that can be processed simultaneously. The role of device performance characteristics is to provide the central control system with a performance boundary model of the physical device, enabling cloud-based planning to accurately predict the operational execution effect without physically accessing the device.

[0028] To transform unstructured device descriptions into computable constraints, the central control system first performs parameter identification and extraction processing on preset device performance characteristics. In this embodiment, the parameter identification process is based on feature classification templates; for example, "Hexa-core CPU" is parsed into a set of clock frequencies for six computing cores, and "H.265 / HEVCdecoding" is mapped to frame rate processing capability at a specific resolution. The extracted set of device target parameters forms a multi-dimensional vector, whose mathematical representation can be constructed as P={p core ,p mem ,p gpu ,p codecThe parameter set is defined as follows: ...}, where each dimension corresponds to a type of performance metric. The resource constraint strategy generated based on this parameter set is defined using a system of inequalities; for example, a memory usage constraint is expressed as... , where m i Let be the memory requirement of the i-th task; the constraint on the number of parallel tasks is: , where c j This indicates the task's occupancy status on the computing cores. Resource constraint policies are generated to translate physical limitations into boundary conditions for the planning algorithm. For example, when planning an 8K video playback task, the resource constraint policy will prohibit the simultaneous scheduling of another 4K video decoding task to avoid frame rate drops due to hardware decoder overload.

[0029] Next, to simulate the entire instruction execution chain in a virtual device environment built in the cloud, the system performs sandboxed simulation processing on the perceived command data according to resource constraint policies. The sandbox environment is implemented using containerization technology, and its resource quotas strictly adhere to the aforementioned constraint policies, such as allocating the same number of shader units to the virtual graphics processor as to the physical machine. The simulation in this embodiment uses a discrete event simulation engine, advancing the execution state step-by-step and dynamically detecting resource conflicts and dependencies. For example, for a composite command like "simultaneously enable face tracking and speech-to-text," the simulation process gradually simulates the computational latency of image frames being passed to the neural network model, the memory usage fluctuations of the speech buffer, and records the load curve of the virtual processor cores. When the simulation detects that the face tracking task requires 90% of the graphics processor resources, while the speech-to-text task requires two CPU cores, the simulation engine compares the available capacity and core count limits of the graphics processor in the resource constraint policy to determine if there is a risk of resource contention. The significance of this process lies in exposing the possibility of deadlock and performance bottlenecks during execution in advance. For example, in the scenario of intelligent digital signage, simulation can reveal memory conflicts between the playback of advertising videos and the launch of interactive games, thereby allowing adjustments to the plan before actual execution.

[0030] After obtaining the pre-execution sequence, the system performs refined resource allocation and timing planning based on device performance characteristics. The resource allocation in this embodiment employs multi-objective optimization based on a genetic algorithm. Its fitness function considers both energy efficiency and execution speed. For example, when allocating video rendering tasks, the system selects the computing power configuration with the lowest power consumption while meeting frame rate requirements, based on the graphics processor energy efficiency curve in the device performance characteristics. Timing planning relies on a directed acyclic graph model, where nodes represent atomic operations and edges represent dependencies. Critical path analysis determines the earliest start time and latest end time of each operation. For operations with hardware sharing conflicts, the system introduces a timing offset strategy. For example, when both screen drawing and camera capture require access to the memory direct access controller, a forced delay δ is inserted into the timing graph based on the memory controller throughput data in the device performance characteristics.t To avoid data contention, the output of time series planning can be formalized as an executable sequence of operations S={(op1...} ,tstart1 ,tend1 ),(op2 ,tstart2 ,tend2 Each operation is assigned an absolute timestamp. These operations enable cross-task collaboration. For example, in a smart home control scenario, commands to gradually brighten lights and open curtains are planned to be executed synchronously. Although they involve different hardware drivers, precise timing alignment creates a smooth scene transition experience.

[0031] Finally, the generated executable operation sequence adopts a multi-layered encoding structure, including an opcode byte stream, a resource descriptor list, and timing metadata. This sequence is encapsulated by a protocol and then sent to the terminal device. Its structural design ensures that the device can execute the operation directly without complex resource decisions. For example, for a complex instruction like "play 4K HDR video and simultaneously adjust ambient light," the executable operation sequence explicitly marks the time window for the video decoding task to exclusively use hardware decoder resources, the time window for the ambient light control task to use the serial peripheral interface bus, and the synchronization triggering conditions for the two operations. By transforming device performance characteristics into computable constraint strategies and then eliminating execution uncertainties through sandbox simulation, an operation sequence that balances efficiency and reliability is ultimately produced, providing a theoretical basis for the collaborative control of heterogeneous device clusters.

[0032] Step S13: Based on the executable operation sequence, perform transmission quality assessment and path selection on the preset propagation path network to obtain the target transmission path.

[0033] The propagation path network, serving as the physical foundation for transmission tasks, is essentially a collection of all available communication links between the central control system and terminal devices. Specifically, it includes heterogeneous network interfaces such as 5G mobile network channels, Wi-Fi wireless links, wired Ethernet connections, and satellite communication channels. The dynamic status parameters of these paths are continuously collected by distributed probe nodes. For example, network quality monitoring services deployed at cloud edge nodes periodically send probe data packets to the target device to measure round-trip latency, bandwidth utilization, packet loss rate, and signal jitter parameters. The propagation path network provides redundant transmission options for remote control commands, ensuring that the system maintains uninterrupted control links even in the event of a single network failure.

[0034] The central control system first traverses all paths in the preset propagation path network, synchronously executing through concurrent probe threads. Each thread extracts a set of real-time status parameters for a specific path. These real-time status parameters encompass both link-layer and network-layer metrics. The latency parameter is calculated using timestamp differentiation, representing the time difference between a data packet being sent from the cloud gateway and the receiving device's response. The bandwidth parameter employs a sliding window measurement method, calculated based on the effective payload successfully transmitted per unit time. The packet loss rate is derived through sequence number continuity analysis, i.e., the proportion of missing sequence numbers in the probe packet sequence. For example, in the control scenario of a smart factory inspection robot, the system simultaneously detects the signal strength of the industrial Wi-Fi channel and the base station load rate of the 5G private network channel; these parameters collectively form the basis for path selection decisions. By traversing and extracting instantaneous snapshots of the transmission environment, dynamic data support is provided for subsequent quality assessment, avoiding decision-making biases caused by using historical network data.

[0035] After obtaining the current status parameters, the system needs to calculate the transmission quality score of each path based on the timing constraint set in the executable operation sequence. The timing constraint set originates from the execution plan generated in step S12 and includes the latest arrival time of each data block and the maximum interval requirement between consecutive data blocks. The quality score calculation used in this embodiment is a multi-objective weighted function, mathematically expressed as Q = w1 × f1 (delay) + w2 × f2 (bandwidth) + w3 × f3 (packet loss rate), where the weight coefficients w1, w2, and f3 are all calculated using the multi-objective weighted function. i Dynamically adjusted based on timing constraints: when the operation sequence includes real-time video stream control, the latency weight is significantly increased; when large-capacity firmware updates are involved, the bandwidth weight takes precedence. Function f1 normalizes the latency parameter, using an sigmoid function to map the original latency value to a score value between 0 and 1; function f2 uses a piecewise linear transformation on the bandwidth parameter to ensure that the score is positively correlated with the actual throughput; function f3 applies a penalty transformation to the packet loss rate, and the score decreases exponentially when the packet loss rate exceeds a threshold. For example, in a telemedicine consultation system, the transmission of ultrasound image data needs to meet the constraint of arriving within 200 milliseconds. In this case, the score for the 5G path will focus on whether its current latency is below 150 milliseconds and retains sufficient margin. The significance of this scoring mechanism is to accurately match network characteristics with business requirements, preventing high-bandwidth but high-jitter paths from being misused in real-time control scenarios.

[0036] The transmission quality scoring system prioritizes all paths in the propagation path network. This process employs an improved topology ranking algorithm that considers not only absolute scores but also the complementary characteristics between paths. The ranking algorithm first establishes a directed graph of path scores, where nodes represent transmission paths and edges represent path switching costs. It generates a transmission execution path sequence by calculating the equilibrium solution between the path with the highest score and the path with the lowest switching cost. During the ranking process, the system intentionally preserves paths with differentiated characteristics. For example, high-latency, high-bandwidth satellite links and low-latency, medium-bandwidth 5G links are grouped into the same priority group, providing alternative options for different types of data blocks. For instance, when autonomous vehicles update high-precision maps, the transmission execution path sequence might prioritize cellular vehicle-to-everything (V2X) channels and secondarily prioritize roadside unit (Roadside Unit) communication, with the two paths complementing each other in the signal coverage area. The significance of generating the transmission execution path sequence lies in constructing a hierarchical transmission scheme, enabling the system to flexibly select transmission strategies based on data importance, ensuring that critical control commands always prioritize high-reliability paths.

[0037] Finally, the target transmission path is selected based on the preset number of paths and the transmission execution path sequence. This decision-making process employs a constraint satisfaction problem-solving model. The preset number of paths is determined by the system's fault tolerance level. For example, a dual-active architecture requires at least two independent paths, while a triple-redundant architecture requires three physically isolated channels. The selection algorithm starts from the top of the transmission execution path sequence and proceeds downwards, simultaneously verifying the fault independence between paths to avoid selecting redundant paths that share the same physical infrastructure. After determining the target transmission path, the system retrieves the corresponding transmission feature data from a preset transmission feature database. This feature data includes protocol encapsulation format, encryption algorithm configuration, forward error correction parameters, and traffic shaping rules. For example, when selecting a millimeter-wave communication path, the transmission feature data specifies the use of a UDP-based proprietary protocol encapsulation to reduce header overhead and enables Reed-Solomon encoding to resist burst errors; while when selecting a wired Ethernet path, the TCP-TLS protocol stack is used to ensure transmission reliability. The purpose of the transmission feature data is to provide specific configuration templates for subsequent protocol encapsulation stages, enabling general operation sequences to adapt to the characteristics of specific transmission media. The significance of the entire path selection mechanism lies in realizing the dynamic adaptation of network resources and control tasks. For example, in the scenario of cross-sea UAV control, the system automatically selects the channel with more stable signal quality between the maritime satellite link and the high-altitude platform base station link to ensure the continuous reach of control commands.

[0038] Step S14: Encapsulate the executable operation sequence according to the target transmission path to obtain a remote control command data stream.

[0039] It should be understood that the transmission feature data obtained in step S13 above, serving as a configuration blueprint for protocol encapsulation, is essentially a derived set of technical parameters for the target transmission path, specifically including two core elements: encapsulation strategy and transmission protocol. The encapsulation strategy defines the organization rules for data units, including fragment size, header field structure, and sequence number encoding scheme; the transmission protocol clarifies the session management mechanism, covering error detection methods, encryption algorithms, and handshake procedures. This feature data is extracted from the transmission feature database, and its content is closely related to the physical characteristics of the path. For example, the encapsulation strategy for high-latency satellite links will enable large data block aggregation, while the transmission protocol for high-jitter mobile networks will employ a fast retransmission mechanism. The role of transmission feature data is to adapt general operation sequences to specific transmission environments, ensuring that control commands maintain optimal transmission efficiency under specific link conditions.

[0040] The central control system first adds transmission sequence tags to each data unit in the executable operation sequence according to the encapsulation strategy. In this embodiment, this operation is implemented through a sequence number allocation engine. The transmission sequence tags are composed of globally monotonically increasing sequence numbers. Each sequence number establishes a mapping relationship with the atomic instructions in the operation sequence, and timestamp information is embedded to mark the logical execution order. During the encapsulation process, the system intelligently segments the large-scale operation data according to the maximum transmission unit of the path. Each segment obtains a sub-sequence number tag and records the parent sequence number association. For example, in the scenario of controlling a robotic arm in a smart factory, the operation sequence containing trajectory coordinates and torque parameters is split into multiple data blocks. The 32-bit sequence number added to the header of each block ensures the reassembly order at the receiving end and provides a basis for packet loss detection. By adding sequence tags to establish the inherent logical order of the data stream, the receiving device can still accurately reconstruct the instruction sequence even when it arrives out of order. Especially in multimedia synchronous control, the temporal correlation between video frames and audio packets is maintained through the tag system.

[0041] Next, the security encryption module performs the addition of checksums and security tokens to the encapsulated data sequence according to the transmission protocol. In this embodiment, the checksum generation uses a cyclic redundancy check (CRC) algorithm, whose polynomial calculation can be expressed as CRC = data mod G(x). The generator polynomial G(x) is selected based on data sensitivity; sensitive control commands use the CRC-32 standard, while ordinary state data uses the CRC-16 standard. The security token in this embodiment is generated based on a two-way authentication framework, containing a combination of a hash value encrypted with the session key and a digital signature. Its structure can be described as Token = Where Ks is the session key and Kpriv is the cloud-based private key. For example, in a medical imaging diagnostic system, operation instructions containing patient privacy data, after being verified with a checksum to ensure integrity, also embed a security token based on elliptic curve cryptography to prevent man-in-the-middle attacks and instruction tampering. This security process constructs an end-to-end trusted transmission channel, defending against transmission errors through checksums and achieving identity authentication and non-repudiation through security tokens, meeting the security compliance requirements of industrial control scenarios.

[0042] After obtaining the encrypted dataset, the system performs sorting and frame assembly operations based on the transmission sequence tags. This process is executed by the protocol assembly engine. The sorting algorithm reconstructs the data unit sequence based on the time sequence information in the tags, using a min-heap data structure to ensure processing efficiency, and rearranging data packets whose order is disordered due to encryption. Frame assembly dynamically adjusts the frame structure according to the characteristics of the transmission path. Low-bandwidth paths use a highly compressed frame format, while high-bandwidth paths use an extended frame structure to carry metadata. For example, in a smart city traffic control system, traffic light status update instructions are assembled into compact data frames, while traffic flow statistics reports use extended frames to carry time dimension information. During the framing process, each frame header contains a start sequence number and a frame length field, enabling the receiver to quickly locate frame boundaries and parse the payload. The significance of this framing assembly lies in optimizing network resource utilization. By designing a frame structure that adapts to path characteristics, it minimizes transmission overhead while ensuring instruction integrity. Especially in IoT device group control scenarios, reasonable framing can significantly reduce the load pressure on base stations.

[0043] Finally, based on the transmission quality score corresponding to the target transmission path, forward error correction coding is added to the control command data stream. The forward error correction parameters are dynamically calculated based on historical path performance data, and their coding strength is negatively correlated with the transmission quality score: strong error correction coding is enabled when the path score is below a threshold, and lightweight verification is used when the score is excellent. The system uses Reed-Solomon coding as the core algorithm, and its coding process can be described as a mathematical transformation mapping k data symbols to n coded symbols (c1). c2 ,...,cn =RSEncode(d1) ,d2 ,...,dk ), where the number of redundant symbols m=n km=n k is dynamically adjusted based on the packet loss rate along the path. For example, in remote control of field exploration equipment, when the wireless link quality score drops below 0.7, the system automatically increases the redundancy of Reed-Solomon coding from 25% to 40%, significantly enhancing its ability to resist sudden interference. Adding forward error correction coding improves the robustness of data transmission. By pre-adding redundant information at the transmitting end, the receiving end can recover the original instructions even if some data is lost or damaged, which is particularly important for continuous data stream services such as real-time video analysis.

[0044] Abstract operation sequences are transformed into concrete network data packets through multi-layered protocol encapsulation. This layered encapsulation architecture not only ensures reliable transmission of control commands in complex network environments, but also achieves an optimal balance between transmission efficiency and reliability through a dynamic adaptation mechanism, providing low-latency, high-fidelity remote control data streams for intelligent multimedia interactive devices.

[0045] The intelligent multimedia interactive device that needs to be controlled (e.g., a smart conference board in a remote conference room, a smart interactive blackboard in a network classroom) includes: a secure communication module that receives remote control command data streams from the cloud, decrypts and verifies their integrity, and reassembles them into an executable operation sequence; encrypts the execution status data generated on the device and sends it back to the cloud; connects to the secure communication module, calls the device's local APIs and drivers to strictly follow the executable operation sequence, and has an instruction execution engine; and monitors the device's own resource status in real time and generates a high-fidelity real-time execution status data resource monitor. The intelligent multimedia interactive device performs the following steps: In the operation of intelligent multimedia interactive devices, the transmission protocol mapping table, serving as a protocol identification dictionary on the device side, is essentially a set of associations between transmission path characteristics and corresponding parsing rules. Specifically, it records the mapping relationship between network port numbers, packet signatures, and protocol stack configurations. This mapping table is securely downloaded from the cloud and stored in a trusted execution environment during device initialization. For example, the firmware of a smart TV obtains the latest protocol mapping table through a secure channel during startup, containing configuration rules for using the QUIC protocol for the Wi-Fi channel and a custom UDP protocol for the 5G channel. When a data stream arrives at the device's network interface, the protocol identification engine first extracts path feature parameters, including the source IP address range, destination port number, and packet time-to-live (TTL) value, and then performs a matching query in the mapping table. For example, when a video conferencing terminal detects that a data stream originates from a specific edge computing node and the destination port is 443, it automatically identifies it as an HTTPS-TLS protocol stream; when the data stream originates from the core network gateway and carries a special differential service code point marker, it is identified as a real-time transport protocol stream. The significance of this path-driven protocol determination mechanism lies in enabling multi-protocol adaptive processing, allowing devices to correctly parse instructions transmitted via different network infrastructures, and ensuring accurate parsing of control instructions delivered via wired Ethernet or wireless private networks in industrial IoT scenarios.

[0046] After the transmission protocol is determined, the intelligent multimedia interactive device performs a deep processing flow of the remote control command data stream. The parsing process peels off the encapsulation header layer by layer according to the protocol specification. First, the network layer and transport layer headers are removed. Then, the application layer encapsulation is processed according to the protocol type: QUIC protocol streams require decryption of the first packet's protective payload, while real-time transmission protocol streams require reordering out-of-order data packets. The verification phase uses a cyclic redundancy check (CRC) code to verify data integrity. Its check polynomial is consistent with that of the cloud encapsulation phase. When verification fails, a retransmission request is triggered instead of directly discarding the data, ensuring the reliability of command transmission. The secure decryption phase uses a session key management module. This module exchanges key materials with the cloud during the device registration phase. The decryption process can be described as P = (C), where The session key is derived through elliptic curve Diffie-Hellman key exchange, where C is the ciphertext payload and P is the restored original data. For example, when a smart security camera decrypts access control commands, it verifies data integrity to prevent command tampering and accelerates decryption operations through a hardware security module, ensuring rapid response to critical security commands. The significance of this entire parsing and verification process lies in building an end-to-end trusted command channel, preventing maliciously forged control commands from triggering abnormal device operations in home automation scenarios, while simultaneously ensuring the instantaneous response capability of low-power devices such as smart locks through efficient decryption.

[0047] After obtaining the reconstructed executable operation sequence, the intelligent multimedia interactive device initiates a categorized execution mechanism based on a partitioning strategy. This partitioning strategy, predefined based on operation type and resource requirements, divides the operation sequence into independent execution domains such as a real-time operation domain, a media processing domain, and a system management domain. The real-time operation domain is dedicated to handling latency-sensitive commands, such as projector focus adjustment or lighting equipment on / off control; this domain employs a preemptive scheduling strategy to ensure timely response. The media processing domain handles audio / video encoding / decoding and graphics rendering tasks, enjoying priority access to the graphics processor and digital signal processor. The system management domain handles background tasks such as firmware updates and log uploads. Each execution domain is equipped with an independent resource monitor, recording parameters such as command start time, end time, resource utilization, and execution status, forming complete execution trajectory data. For example, when a digital signage device executes a video switching command, the media domain records the decoder initialization time and frame rendering latency, while the system domain synchronously records the number of read / write operations on the storage device. This trajectory data collectively constitutes a panoramic view of the device's status. The significance of categorized execution lies in achieving resource isolation and priority assurance. In smart classroom scenarios, it ensures that screen writing commands triggered by the teacher take precedence over content synchronization requests from the student, maintaining a smooth teaching and interactive experience.

[0048] Ultimately, the intelligent multimedia interactive device transmits the generated real-time execution status data back to the central control system via a multi-level aggregated transmission path. This multi-level aggregated transmission path is a hierarchical transmission mechanism designed based on the criticality of the status data. Critical status data, such as device anomaly alarms or security events, are directly transmitted to the cloud via a high-priority path; performance indicator data, such as processor load and memory usage, are uploaded periodically via a standard path; and historical operation logs are transmitted via a batch processing path during network downtime. Before transmission, the status data undergoes compression and serialization. Numerical status data uses differential encoding to reduce data volume, while enumerated status data uses dictionary encoding to eliminate redundancy. The transmission path selector dynamically evaluates the current quality of each path and allocates dedicated transmission channels for emergency status data. For example, when a medical monitoring device detects abnormal patient vital signs, it automatically activates parallel transmission via both satellite communication links and mobile networks to ensure reliable delivery of alarm information. During encapsulation, device identification and data classification tags are added to the status data packets, enabling the cloud to quickly route them to the corresponding virtual device image for status updates. The significance of this multi-level aggregation transmission mechanism lies in optimizing uplink bandwidth utilization. In smart factory scenarios, it enables the status data of hundreds of sensor nodes to be synchronized to the cloud in real time without blocking the control channel, providing continuous data input for the digital twin of the production system.

[0049] The complete process, from protocol recognition to status feedback, constructs a closed-loop two-way communication system between the device and the cloud. Precise protocol adaptation ensures reliable execution of control commands, partitioned resource management guarantees real-time operational responses, and an intelligent status feedback mechanism synchronizes system status updates. Through these operations, intelligent multimedia interactive devices, under a centralized cloud control architecture, can accurately execute complex operational commands and promptly report their own status changes. This provides a data foundation for the dynamic optimization of remote control systems, enabling refined management and collaborative control of large-scale device clusters in smart city and industrial automation scenarios.

[0050] Furthermore, after issuing commands, the central control system enters the status feedback processing stage, with the core objective of achieving system self-optimization. Real-time execution status data, serving as the carrier of the feedback loop, is essentially a complete record of the command execution process by the intelligent multimedia interactive device, including structured information such as operation start timestamp, end timestamp, actual resource consumption, execution result code, and abnormal event records. This data is transmitted to the cloud via a multi-level aggregation transmission path at the device end. During transmission, differential compression and entropy coding techniques are used to reduce network load. For example, continuously collected processor utilization sequences are converted into key point sets using linear interpolation algorithms, and recurring error codes are replaced with dictionary indices. The role of real-time execution status data is to provide the cloud with a true reflection of device operation, enabling the central control system to go beyond static device parameters and dynamically optimize control strategies based on actual performance.

[0051] The central control system initiates a decompression process on the received real-time execution status data. This process employs a reverse decoding pipeline corresponding to the compression algorithm used on the device side. The compressed data packet first undergoes frame structure parsing, separating the header metadata from the payload. The header information indicates the compression algorithm version and data segmentation strategy used. For numerical data using differential encoding, the decompression engine reconstructs the original data sequence using base values ​​and difference sequences. For enumerated data using dictionary encoding, the decompression module restores the compact index to a complete status description by querying the dictionary table synchronized on the device side. For example, in a smart retail terminal system, after decompression, the sales data packet returned by the device restores the complete transaction time series, inventory change records, and a set of device anomaly events. The significance of decompression processing lies in reconstructing a readable status information system, providing a structured data foundation for subsequent analysis, and ensuring the integrity and temporal continuity of vital sign data are not affected by transmission compression in smart medical device monitoring.

[0052] After obtaining the execution feedback data, the system initiates a calibration and update process for the device performance characteristics through a preset dynamic adjustment method. This dynamic adjustment method is based on an adaptive mechanism in control theory, comparing the execution feedback data with the predicted values ​​of the virtual device image to establish an error propagation model. The calibration process first identifies the direction of deviation in performance characteristic parameters. For example, when the actual time consumed by a graphics rendering task consistently exceeds the predicted value by 20%, the graphics processor's computing power parameter is flagged as needing to be lowered; when the actual completion time of a memory-intensive task is consistently shorter than expected, the memory bandwidth parameter is determined to be underestimated. In this embodiment, the parameter update uses a recursive least squares method with a forgetting factor. The significance of this dynamic adjustment is to ensure that the virtual device image continuously approximates the real capabilities of the physical device. In autonomous driving test scenarios, when the vehicle control system frequently reports unexpected steering delays, the cloud will adjust the control response parameters of the vehicle model accordingly to ensure the accuracy of the simulation test.

[0053] The updating of equipment performance characteristics follows a gradual adjustment principle to avoid drastic parameter fluctuations caused by a single abnormal feedback. The system sets confidence intervals and update step size limits for each type of performance parameter. For example, the adjustment range for processor clock speed parameters is limited to 5% each time, and corrections to memory capacity parameters require three consecutive feedback data points supporting a change in the same direction. Updated equipment performance characteristics take effect immediately in the new planning cycle, but the system retains historical versions as rollback backups. For instance, in a large digital signage network, when a certain model of equipment is detected to have reduced decoding performance under high temperatures, the system dynamically adds a temperature-performance correlation model to the virtual device image, enabling subsequent video content planning to automatically avoid highly complex encoding formats. The significance of the dynamic calibration mechanism lies in building a control system with learning capabilities, adapting to the natural degradation of equipment performance and changes in operating conditions in the industrial IoT environment, and ensuring consistent production quality through continuous optimization.

[0054] The aforementioned feedback and adjustment process constitutes a self-evolutionary loop for the cloud-based control system, feeding back the actual performance of the equipment to the planning model and continuously narrowing the gap between the virtual environment and the physical world. This data-driven dynamic adjustment mechanism enables the remote control system to cope with equipment heterogeneity and performance fluctuations. In large-scale deployment scenarios such as smart homes and industrial automation, it continuously learns and optimizes the global control strategy, ultimately achieving a dual improvement in system reliability and control accuracy. The entire feedback processing flow demonstrates the core advantages of a centralized cloud architecture, transforming distributed equipment operating experience into centralized intelligent evolutionary power, providing a continuous optimization technical foundation for collaborative equipment control in complex environments.

[0055] Please see Figure 4 , Figure 4 This is a schematic diagram of the structure of a remote control device for an intelligent multimedia interactive device provided in an embodiment of this application. Figure 4 As shown, the remote control device 2 of the intelligent multimedia interactive device includes: an interaction parsing module 21, a planning and allocation module 22, a path selection module 23, a protocol encapsulation module 24, and a feedback adjustment module 25.

[0056] The interaction parsing module 21 is used to perform layered parsing processing on the user-input interaction data to generate perception command data; The planning and allocation module 22 is used to perform execution planning and resource allocation on the sensing command data according to the preset device performance characteristics, and generate an executable operation sequence; The path selection module 23 is used to perform transmission quality assessment and path selection on the preset propagation path network according to the executable operation sequence, so as to obtain the target transmission path. Protocol encapsulation module 24 is used to encapsulate the executable operation sequence according to the target transmission path to obtain a remote control instruction data stream.

[0057] In one possible implementation, the interaction parsing module 21 is specifically used for: Based on the preset interaction control type, the user input interaction data is classified and processed to obtain first-level interaction data and second-level interaction data; According to the preset command mapping strategy, the first-level interactive data is converted into the first-level command; The second-level interactive data is semantically parsed and intent is identified to generate the second-level command.

[0058] In one possible implementation, the planning and allocation module 22 is specifically used for: The preset parameters of the preset equipment performance characteristics are identified and extracted to obtain the set of target parameters of the equipment, and the resource constraint strategy of the corresponding equipment is generated according to the set of target parameters of the equipment. The perception command data is subjected to sandbox simulation processing based on the resource constraint strategy to obtain a pre-execution sequence. Based on the device performance characteristics, resource allocation and timing planning are performed on each operation step in the pre-execution sequence to generate an executable operation sequence.

[0059] In one possible implementation, the path selection module 23 is specifically used for: Traverse each path in the preset propagation path network to extract the current state parameters corresponding to each path; Calculate the transmission quality score of the corresponding path based on the set of timing constraints in the executable operation sequence and the current state parameters; Based on the transmission quality score, the paths in the propagation path network are prioritized for transmission, and a transmission execution path sequence is generated. The target transmission path is selected from the propagation path network according to the preset number of paths and the transmission execution path sequence, and the corresponding transmission feature data is obtained from the preset transmission feature database according to the target transmission path.

[0060] In one possible implementation, the protocol encapsulation module 24 is specifically used for: According to the encapsulation strategy, a transmission order tag is added to each data in the executable operation sequence to obtain an encapsulated data sequence; According to the transmission protocol, a checksum and a security token are added to each piece of data in the encapsulated data sequence to generate an encrypted dataset; Based on the transmission order tags corresponding to each data in the encrypted dataset, the encrypted dataset is sorted and framed to generate a control command data stream. Based on the transmission quality score corresponding to the target transmission path, forward error correction coding is added to each data frame in the control command data stream to generate a remote control command data stream.

[0061] In one possible implementation, the remote control device 2 of the intelligent multimedia interactive device further includes a feedback adjustment module 25, which is used for: The received real-time execution status data from the intelligent multimedia interactive device is decompressed to obtain execution feedback data; The device performance characteristics are dynamically calibrated and updated based on the execution feedback data using a preset dynamic adjustment method.

[0062] The interactive parsing module 21, planning and allocation module 22, path selection module 23, protocol encapsulation module 24, and feedback adjustment module 25 can all be implemented in software or hardware. For example, the implementation of the interactive parsing module 21 will be described below. Similarly, the implementation of the planning and allocation module 22, path selection module 23, protocol encapsulation module 24, and feedback adjustment module 25 can refer to the implementation of the interactive parsing module 21.

[0063] As an example of a software functional unit, the interaction resolution module 21 may include code running on a computing instance. The computing instance may include at least one of a physical host (computing device), a virtual machine, or a container. Further, the aforementioned computing instance may be one or more. For example, the interaction resolution module 21 may include code running on multiple hosts / virtual machines / containers. It should be noted that the multiple hosts / virtual machines / containers used to run the code may be distributed in the same region or in different regions. Further, the multiple hosts / virtual machines / containers used to run the code may be distributed in the same availability zone (AZ) or in different AZs, each AZ including one or more geographically proximate data centers. Typically, a region may include multiple AZs.

[0064] Similarly, multiple hosts / virtual machines / containers used to run this code can be distributed within the same Virtual Private Cloud (VPC) or across multiple VPCs. Typically, a VPC is set up within a region. Communication between two VPCs within the same region, as well as between VPCs in different regions, requires a communication gateway to be set up within each VPC to enable interconnection between VPCs.

[0065] As an example of a hardware functional unit, the interactive parsing module 21 may include at least one computing device, such as a server. Alternatively, the interactive parsing module 21 may also be a device implemented using an application-specific integrated circuit (ASIC) or a programmable logic device (PLD). The PLD may be implemented using a complex programmable logical device (CPLD), a field-programmable gate array (FPGA), generic array logic (GAL), or any combination thereof.

[0066] The interactive resolution module 21 includes multiple computing devices that can be distributed within the same region or in different regions. Similarly, the interactive resolution module 21 includes multiple computing devices that can be distributed within the same Availability Zone (AZ) or in different AZs. Likewise, the interactive resolution module 21 includes multiple computing devices that can be distributed within the same Virtual Private Cloud (VPC) or in multiple VPCs. These multiple computing devices can be any combination of computing devices such as servers, ASICs, PLDs, CPLDs, FPGAs, and GALs.

[0067] See Figure 5 As shown, Figure 5 This is a schematic diagram of a computing device provided in this application. The computing device 100 includes: a processor 104, a communication interface 108, a bus 102, and a memory 106. The processor 104, the communication interface 108, and the memory 106 communicate via the bus 102. In practical applications, communication can also be achieved through other means such as wireless transmission; however, this is not limited here.

[0068] The computing device 100 may be a server or a terminal device. It should be understood that this application does not limit the number of processors and memory in the computing device 100.

[0069] The processor 104 may include any one or more processors such as a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP).

[0070] The communication interface 108 uses transceiver modules such as, but not limited to, network interface cards and transceivers to enable communication between the computing device 100 and other devices or communication networks.

[0071] Bus 102 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 5 The bus 102 may be represented by a single line, but this does not mean that there is only one bus or one type of bus. The bus 102 may include a path for transmitting information between various components of the computing device 100 (e.g., memory 106, processor 104, communication interface 108).

[0072] Memory 106 may include volatile memory, such as random access memory (RAM). Memory 106 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).

[0073] The memory 106 stores executable program code, and the processor 104 executes the executable program code to implement the functions of the aforementioned interactive parsing module 21, planning and allocation module 22, path selection module 23, protocol encapsulation module 24 and feedback adjustment module 25 respectively, thereby realizing the remote control method of the intelligent multimedia interactive device. That is, the memory 106 stores instructions for executing the remote control method of the intelligent multimedia interactive device.

[0074] This application also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that a computing device can store, or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive). The computer-readable storage medium includes instructions that instruct the computing device to perform a remote control method for an intelligent multimedia interactive device, or instruct the computing device to perform a remote control method for an intelligent multimedia interactive device.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention 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; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. A remote control method for an intelligent multimedia interactive device, applied to a central control system, characterized in that, The method includes: The user-input interaction data is processed in layers to generate perception command data; Based on the preset device performance characteristics, the perception command data is executed and resources are allocated to generate an executable operation sequence; Based on the executable operation sequence, the transmission quality of the preset propagation path network is evaluated and the path is selected to obtain the target transmission path; The executable operation sequence is protocol-encapsulated according to the target transmission path to obtain a remote control command data stream.

2. The remote control method for an intelligent multimedia interactive device according to claim 1, wherein the sensing command data includes a first-level command and a second-level command, characterized in that, The process of performing layered parsing of user-input interaction data to generate perception command data includes: Based on the preset interaction control type, the user input interaction data is classified and processed to obtain first-level interaction data and second-level interaction data; According to the preset command mapping strategy, the first-level interactive data is converted into the first-level command; The second-level interactive data is semantically parsed and intent is identified to generate the second-level command.

3. The remote control method for an intelligent multimedia interactive device according to claim 1, characterized in that, The step of performing execution planning and resource allocation on the perceived command data based on preset device performance characteristics to generate an executable operation sequence includes: The preset parameters of the preset equipment performance characteristics are identified and extracted to obtain the set of target parameters of the equipment, and the resource constraint strategy of the corresponding equipment is generated according to the set of target parameters of the equipment. The perception command data is subjected to sandbox simulation processing based on the resource constraint strategy to obtain a pre-execution sequence. Based on the device performance characteristics, resource allocation and timing planning are performed on each operation step in the pre-execution sequence to generate an executable operation sequence.

4. The remote control method for an intelligent multimedia interactive device according to claim 1, characterized in that, The step of evaluating the transmission quality and selecting a path for a preset propagation path network based on the executable operation sequence to obtain the target transmission path includes: Traverse each path in the preset propagation path network to extract the current state parameters corresponding to each path; Calculate the transmission quality score of the corresponding path based on the set of timing constraints in the executable operation sequence and the current state parameters; Based on the transmission quality score, the paths in the propagation path network are prioritized for transmission, and a transmission execution path sequence is generated. The target transmission path is selected from the propagation path network according to the preset number of paths and the transmission execution path sequence, and the corresponding transmission feature data is obtained from the preset transmission feature database according to the target transmission path.

5. The remote control method for an intelligent multimedia interactive device according to claim 4, wherein the transmission feature data includes an encapsulation strategy and a transmission protocol, characterized in that, The step of protocol encapsulating the executable operation sequence according to the target transmission path to obtain a remote control command data stream includes: According to the encapsulation strategy, a transmission order tag is added to each data in the executable operation sequence to obtain an encapsulated data sequence; According to the transmission protocol, a checksum and a security token are added to each piece of data in the encapsulated data sequence to generate an encrypted dataset; Based on the transmission order tags corresponding to each data in the encrypted dataset, the encrypted dataset is sorted and framed to generate a control command data stream. Based on the transmission quality score corresponding to the target transmission path, forward error correction coding is added to each data frame in the control command data stream to generate a remote control command data stream.

6. The remote control method for an intelligent multimedia interactive device according to claim 1, characterized in that, The method further includes: The received real-time execution status data from the intelligent multimedia interactive device is decompressed to obtain execution feedback data; The device performance characteristics are dynamically calibrated and updated based on the execution feedback data using a preset dynamic adjustment method.

7. A remote control device for an intelligent multimedia interactive device, applied to the remote control method of the intelligent multimedia interactive device as described in claim 1, characterized in that, The device includes: The interaction parsing module is used to perform layered parsing and processing of user-input interaction data to generate perception command data; The planning and allocation module is used to perform execution planning and resource allocation on the sensing command data according to preset device performance characteristics, and generate an executable operation sequence. The path selection module is used to perform transmission quality assessment and path selection on the preset propagation path network according to the executable operation sequence, so as to obtain the target transmission path. The protocol encapsulation module is used to encapsulate the executable operation sequence according to the target transmission path to obtain a remote control command data stream.

8. A computing device, characterized in that, The computing device includes: At least one processor; and, A memory and a communication interface that are communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the at least one processor implements the remote control method of the intelligent multimedia interactive device according to any one of claims 1 to 6 by executing the instructions stored in the memory.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the remote control method for the intelligent multimedia interactive device according to any one of claims 1 to 6.

Images