A multimedia device control method and system based on adaptive protocol matching

Abstract

The application relates to a multimedia device control method and system based on adaptive protocol matching. The method comprises the following steps: firstly, collecting communication protocol data to extract features to form a protocol feature data set. Secondly, confidence checking is performed on the data set type identification result, and if the result is lower than a threshold value, deep features are extracted through a convolutional neural network; based on the classification result, a national standard protocol adaptation conversion engine is called to convert heterogeneous protocols into unified format data, and the data is encrypted through a national encryption algorithm and two-way identity verification to generate a secure communication data frame. Finally, hash checking and structured analysis are performed on the data frame, interactive information packets are constructed in combination with user behavior characteristics, control strategies are dynamically optimized through real-time feedback data, and an adaptive interactive control scheme is generated. Through the technical closed loop of "intelligent analysis-standard conversion-safety reinforcement-closed loop optimization", the efficiency, intelligent level and safety performance of multimedia device management are significantly improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of computer application, and particularly relates to a multimedia device control method and system based on adaptive protocol matching. BACKGROUND

[0002] In the information-based multimedia teaching and management scene, the protocol compatibility problem caused by the heterogeneity of multimedia devices has become an industry pain point. Different manufacturers' central control, projection, sound equipment and the like generally adopt differentiated communication protocols (such as Modbus RTU, self-defined TCP protocol and the like), so that the devices cannot be directly connected to a unified management platform and need to rely on manual protocol adaptation, resulting in high deployment cost and low operation and maintenance efficiency. At the same time, the traditional control scheme lacks dynamic adaptive capability, is difficult to be compatible with new device protocols, and the communication security protection means is single, and cannot meet the strict requirements of the education industry on data security (such as the third level standard of equal protection 2.0) and real-time remote control. With the large-scale application of Internet of Things devices, how to realize the "plug and play" of multi-manufacturer devices, automatic protocol analysis and conversion and secure and reliable communication has become a technical problem to be solved. SUMMARY

[0003] Therefore, it is necessary to provide a multimedia device control method and system based on adaptive protocol matching, which can significantly improve the efficiency and security of multimedia device management.

[0004] In a first aspect, the application provides a multimedia device control method based on adaptive protocol matching, comprising:

[0005] Obtaining communication protocol data of a multimedia device, constructing a feature vector space and comparing with a preset protocol feature library, determining protocol type, version information and manufacturer private protocol features, and obtaining protocol feature data set.

[0006] Performing protocol type identification on the protocol feature data set, and if the identification result is lower than a preset confidence threshold, performing deep feature extraction to obtain protocol classification data.

[0007] Based on the protocol classification data, calling a national standard protocol adaptation conversion engine, converting the data into a unified format, loading a national secret level encryption algorithm and a two-way identity verification rule, and generating a secure communication data frame.

[0008] Performing hash check on the secure communication data frame, and parsing the data stream that passes the integrity verification into structured interactive information, combining user historical operation records to extract behavior habit features, and generating a preliminary interactive information package.

[0009] According to the real-time user operation feedback and the equipment running state data of the unified platform, the interactive information package is judged, and if it does not meet the current control demand, an interactive control scheme is generated through model optimization information.

[0010] In one of the embodiments, the communication protocol data of the multimedia device is acquired, a feature vector space is constructed, and a preset protocol feature library is compared to determine protocol type, version information, and manufacturer private protocol features to obtain a protocol feature dataset, including:

[0011] The original message data of the multimedia device communication protocol is acquired; the original message data contains binary streams in the protocol interaction process.

[0012] The original message data is preprocessed, and key field sequences are extracted according to the preprocessed protocol message; the key field sequences include protocol header identification, frame header identification, function code, data bit length, and check mode.

[0013] The key field sequences are converted into multi-dimensional feature vectors, and a protocol classification model is constructed based on a gradient boosting model combined with a Bayesian optimization algorithm; the multi-dimensional feature vectors include field distribution entropy values, jump frequency parameters, and weighted feature indicators.

[0014] Based on the trained protocol classification model, the cosine similarity algorithm is used to calculate the similarity matrix of the multi-dimensional feature vectors and the preset protocol feature library; the preset protocol feature library stores standard feature vectors of known protocol types.

[0015] The protocol type and version number are matched according to the similarity matrix; the version number corresponds to the maximum similarity entry in the preset protocol feature library.

[0016] The unmatched protocol feature fields are extracted as private protocol features; the private protocol features include specific manufacturer-defined check bit patterns and custom instruction codes.

[0017] The protocol type, version number, and private protocol features are written into the preset protocol feature library to obtain the protocol feature dataset.

[0018] In one of the embodiments, the protocol type of the protocol feature dataset is identified, and if the identification result is lower than the preset confidence threshold, deep feature extraction is performed to obtain protocol classification data, including:

[0019] The multi-dimensional feature vectors in the protocol feature dataset are input into the preset classification model, and the protocol type identification result and the corresponding confidence value are output.

[0020] It is judged whether the confidence value is lower than the preset confidence threshold; if the confidence value is lower than the confidence threshold, the multi-dimensional feature vectors are processed again using a convolutional neural network to extract protocol deep feature vectors.

[0021] According to the protocol depth feature vector, a protocol classification template in a preset protocol feature library is matched to obtain a protocol classification label.

[0022] Protocol classification data is generated based on the protocol classification label; the protocol classification data includes a protocol type identifier, a confidence parameter, and a private feature label.

[0023] In one embodiment, a national standard protocol adaptation conversion engine is called based on the protocol classification data, a unified format data is converted, national secret level encryption algorithms and bidirectional identity verification rules are loaded, and a secure communication data frame is generated, including:

[0024] According to the source protocol identifier in the protocol classification data, a protocol analysis template in the national standard protocol adaptation conversion engine is matched to generate intermediate structure data.

[0025] The intermediate structure data is standardized in data format and reorganized according to a preset field order and type to obtain unified format data.

[0026] According to the data sensitivity level of the unified format data, a national secret algorithm type is selected, and a key factor is injected to generate an encryption parameter.

[0027] The encryption parameter is bound with an identity certificate in a bidirectional verification strategy to construct a verification data packet containing a timestamp, a random number, and a device fingerprint.

[0028] Through the bidirectional verification strategy, an asymmetric encryption operation is performed on the verification data packet to generate a secure communication data frame carrying a digital signature.

[0029] In one embodiment, a hash check is performed on the secure communication data frame, and a data stream that passes the integrity verification is parsed into structured interaction information, behavior habit features are extracted from user historical operation records to generate preliminary interaction information packets, including:

[0030] A hash check is performed on the secure communication data frame using a preset hash algorithm to generate a check result.

[0031] According to the check result, the data integrity is judged, and if the check passes, the data stream is disassembled into an operation instruction field and a timestamp sequence according to a preset interaction protocol template.

[0032] Operation frequency parameters and instruction type distribution data in the user historical operation records are extracted, and operation interval feature values are calculated in combination with the timestamp sequence.

[0033] The operation instruction field and the operation interval feature value are input into a hidden Markov model to output dynamic weight parameters and behavior pattern labels.

[0034] According to the dynamic weight parameters, the parsing priority of the interaction protocol template is adjusted to generate an interaction information packet carrying the behavior pattern label.

[0035] In one embodiment, the interactive information package is judged according to real-time user operation feedback and equipment running state data of the unified platform, and if it does not meet the current control demand, an interactive control scheme is generated through model optimization information, including:

[0036] Real-time user operation feedback data and equipment running state data are obtained; the data includes a user behavior feature vector and an equipment state index matrix.

[0037] The user behavior feature vector is calculated for compliance according to a preset demand matching rule, and a compliance label of the interactive information package is obtained.

[0038] A dynamic weight matrix is constructed according to the correlation degree of the compliance label and the equipment state index matrix.

[0039] When the eigenvalue in the dynamic weight matrix exceeds a preset optimization trigger threshold, an abnormal fluctuation mode in the equipment state data is extracted.

[0040] The abnormal fluctuation mode is input into a pre-trained proximal policy optimization model to generate a model parameter increment set for the current control scene.

[0041] The original control strategy is updated online based on the model parameter increment set to obtain an interactive control scheme containing an adaptive adjustment coefficient.

[0042] In one embodiment, the dynamic weight matrix is constructed by the following formula:

[0043] ;

[0044] ;

[0045] ;

[0046] wherein, denotes the normalized dynamic weight matrix, denotes the dynamic weight matrix, denotes the initial weight vector, , denotes the real-time adjustment matrix, =diag( ), denotes the i th real-time adjustment coefficient, denotes the historical mean of the i th index, denotes the standard deviation of the i th index, ​​​​​represents an adjustment intensity coefficient, represents an anomaly threshold coefficient.

[0047] In a second aspect, the present application also provides a multimedia device control system based on adaptive protocol matching, the system comprising:

[0048] A protocol feature analysis module is configured to acquire communication protocol data of the multimedia device, compare a feature vector space with a preset protocol feature library, determine protocol type, version information and manufacturer private protocol features, and obtain protocol feature data sets; and is further configured to perform protocol type identification on the protocol feature data sets, and if the identification result is lower than a preset confidence threshold, perform deep feature extraction to obtain protocol classification data.

[0049] A protocol security reinforcement module is configured to call a national standard protocol adaptive conversion engine based on the protocol classification data, convert the protocol classification data into unified format data, load national secret level encryption algorithms and bidirectional identity verification rules, and generate secure communication data frames.

[0050] An interactive information processing module is configured to perform hash check on the secure communication data frames, parse data streams that pass integrity verification into structured interactive information, extract behavior habit features in combination with user historical operation records, and generate preliminary interactive information packages.

[0051] A feedback control optimization module is configured to judge the interactive information packages according to real-time user operation feedback and device running state data of a unified platform, and if the interactive information packages do not meet current control requirements, generate interactive control schemes through model optimization information.

[0052] In a third aspect, the present application also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the foregoing method when executing the computer program.

[0053] In a fourth aspect, the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the foregoing method.

[0054] The multimedia device control method, system, computer device and storage medium based on adaptive protocol matching have the following advantages. First, the communication protocol data of the device is collected, the key fields such as the frame header identifier and the function code are extracted and converted into a multi-dimensional feature vector, the gradient boosting model and the cosine similarity algorithm are used to complete the identification and classification of the protocol type, version and manufacturer private characteristics, and the protocol characteristic data set is formed. Second, the confidence of the type identification result of the data set is checked, if the confidence is lower than the threshold, the deep features are extracted through the convolutional neural network, the national standard protocol conversion engine is called based on the classification result, the heterogeneous protocol is converted into the unified format data, the national secret algorithm is used for encryption, the two-way identity authentication is performed to generate the secure communication data frame. Finally, the hash check and the structured analysis are performed on the data frame, the interactive information package is constructed in combination with the user behavior characteristics, the control strategy is dynamically optimized through the real-time feedback data, and the adaptive interactive control scheme is generated. Through the technical closed loop of 'intelligent analysis-standard conversion-safety reinforcement-closed loop optimization', the intelligent level and the safety performance of the multimedia device management are significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the embodiments or the related art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating any creative labor.

[0056] Figure 1 A flow chart of a multimedia device control method based on adaptive protocol matching provided by an embodiment of the present application is provided.

[0057] Figure 2 A structural block diagram of a multimedia device control system based on adaptive protocol matching provided by an embodiment of the present application is provided. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0059] In one embodiment, as shown in Figure 1 A multimedia device control method based on adaptive protocol matching is provided, which can include the following steps:

[0060] In step S101, the communication protocol data of the multimedia device is acquired, the feature vector space is constructed, and the pre-set protocol characteristic library is compared to determine the protocol type, version information and manufacturer private protocol characteristics, and the protocol characteristic data set is obtained.

[0061] Specifically, through RS-232 / 485, TCP / IP and other multi-communication interfaces, the binary original communication protocol data generated by multimedia devices (such as central control and projection) in the interactive process is captured in real time. After preprocessing of the original communication protocol data such as noise reduction and field segmentation, key protocol fields such as frame header identification, function code, data bit length and check mode are extracted to form a structured field sequence. The N-gram sequence analysis algorithm is used to analyze the instruction syntax characteristics, and the importance of each field is calculated by combining the weight model (such as frame header identification weight ratio 10%), and the field sequence is converted into a multi-dimensional feature vector containing field distribution entropy value, jump frequency parameter and weighted feature index. Through the cosine similarity algorithm, the vector is compared with the pre-trained preset protocol feature library (storing the standard feature vector of known protocols such as Modbus RTU / TCP), the protocol type, version number (such as Modbus RTU v1.5) and manufacturer private characteristics (such as CRC16 check mode unique to a manufacturer) are matched, and finally the protocol feature data set containing the protocol type, version and private characteristics is generated.

[0062] Step S102, protocol type identification is performed on the protocol feature data set, and if the identification result is lower than the pre-set confidence threshold, deep feature extraction is performed to obtain protocol classification data.

[0063] Specifically, the multi-dimensional feature vector in the protocol feature data set is input into the classification engine based on the LightGBM gradient boosting model, and the protocol type identification result (such as “custom TCP protocol”) and the corresponding confidence value (probability value, range 0-1) are output. The pre-set confidence threshold (such as 0.7) is used to judge the reliability of the identification result: if the confidence value is lower than the threshold, it means that the current feature information is not enough for accurate classification, triggering the deep feature extraction process—using convolutional neural network (CNN) to process the original feature vector again, extracting deep features such as context semantic association and protocol instruction time sequence dependence, forming a protocol deep feature vector. By matching the classification template (such as the national standard protocol template) in the preset protocol feature library again, a more accurate protocol classification label (such as “manufacturer A custom protocol V2.0”) is obtained, and protocol classification data containing type identification, confidence parameter and private feature label is generated.

[0064] Step S103, based on the protocol classification data, the national standard protocol adaptation conversion engine is called to convert the unified format data and load the national secret level encryption algorithm and the two-way identity verification rule to generate secure communication data frames.

[0065] According to the source protocol identification (such as "manufacturer B's private protocol") in the protocol classification data, the national standard protocol adaptation conversion engine is called, based on GB / T32960 and other national standards or education industry specifications, the instruction semantics (such as the manufacturer's self-defined "0x01" instruction) of the heterogeneous protocol is mapped to the unified semantics (such as "device startup") defined by the national standard through the rule engine, and the intermediate structure data (such as JSON format) is generated. After the intermediate data is standardized in format (such as field order, data type unification), according to the data sensitivity level (public / secret / top secret), the national secret algorithm (SM2 / SM3 / SM4) is automatically selected, the key factor (such as 256-bit key) is injected to generate the encryption parameter. The encryption parameter is bound with the device certificate, user Token and other identity credentials to construct a verification data packet containing timestamp, random number, device fingerprint, and the two-way identity verification is performed through the SM2 asymmetric encryption algorithm, and finally the secure communication data frame carrying the digital signature is generated, which meets the Level 3 security standard of Cybersecurity 2.0.

[0066] Step S104, performing hash check on the secure communication data frame, and parsing the data stream that passes the integrity verification into structured interaction information, extracting behavior habit features from user historical operation records to generate preliminary interaction information package.

[0067] Multi-level hash check (SHA-256 block check+CRC32 cyclic redundancy check) is performed on the secure communication data frame, and the hash values of the sending end and the receiving end are compared to verify the data integrity. If the check passes, according to the pre-set interaction protocol template (such as RESTful API format), the data stream is disassembled into operation instruction field (such as "volume adjustment +5dB") and timestamp sequence. Extract the operation frequency parameter (such as "projection switch" daily call 20 times) and instruction type distribution (such as 70% device control instruction) in the user historical operation record, and calculate the interval characteristic value (such as average response time, periodicity intensity) based on the timestamp. Input the instruction field and interval characteristic value into the hidden Markov model (HMM), predict the user's intention and generate the behavior pattern label (such as "meeting mode" "energy saving mode"), and output the dynamic weight parameter (such as high frequency instruction priority promotion 30%) at the same time, finally adjust the parsing order of the interaction protocol template, and generate the preliminary interaction information package carrying the behavior label.

[0068] Step S105, according to the real-time user operation feedback and device running state data of the unified platform, the interaction information package is judged, if it does not meet the current control demand, the model optimization information is generated to generate the interaction control scheme.

[0069] Specifically, by collecting user operation feedback data (such as interface clicks, instruction execution results) and device state indicators (throughput, error rate, response time) in real time through the unified management platform, a user behavior feature vector (such as operation sequence encoding) and a device state indicator matrix are constructed. The behavior feature vector is calculated for compliance (such as instruction permission verification, parameter range verification) using a demand matching rule engine, and a compliance label (pass / warning / reject) is generated. Based on the Pearson correlation coefficient, the correlation between the compliance label and the device indicators is calculated, and a dynamic weight matrix is constructed. The weight of abnormal indicators (such as error rate exceeding the mean value + 1.5 times the standard deviation) is dynamically increased (such as from 20% to 40%). When the eigenvalue of the weight matrix exceeds the optimization trigger threshold, the abnormal fluctuation pattern in the device state (such as response time continuously for 5 minutes > 100 ms) is extracted, and a near-end policy optimization (PPO) model is used to generate a parameter increment set (such as adjusting the thread pool size, cache strategy), which is used to update the original control strategy online, and finally an interactive control scheme containing adaptive adjustment coefficients is generated, realizing real-time optimization and closed-loop feedback of control logic.

[0070] The above-mentioned multimedia device control method based on adaptive protocol matching first extracts frame header identification, function code and other key fields from device communication protocol data and converts them into a multi-dimensional feature vector. The gradient boosting model and cosine similarity algorithm are used to identify and classify the protocol type, version and manufacturer's private features, forming a protocol feature data set. Second, the confidence of the data set type identification result is checked. If it is below the threshold, the deep features are extracted through the convolutional neural network. Based on the classification result, the national standard protocol adaptation conversion engine is called to convert the heterogeneous protocol into a unified format data, and the national secret algorithm is used for encryption and two-way identity authentication to generate secure communication data frames. Finally, the data frames are subjected to hash check and structured analysis, and the interactive information package is constructed in combination with the user behavior features. The control strategy is dynamically optimized through real-time feedback data, and an adaptive interactive control scheme is generated. This method realizes the technical closed loop of "intelligent analysis-standard conversion-safety reinforcement-closed loop optimization", which significantly improves the intelligent level and safety performance of multimedia device management.

[0071] In one embodiment, the communication protocol data of the multimedia device is obtained, the feature vector space is constructed, and the preset protocol feature library is compared to determine the protocol type, version information and manufacturer's private protocol features, and a protocol feature data set is obtained, which can include the following steps:

[0072] Step S201, obtaining multimedia device communication protocol raw message data; the raw message data contains binary stream in the protocol interaction process.

[0073] Step S202, the original message data is preprocessed, and key field sequences are extracted according to the preprocessed protocol message; the key field sequences include protocol header identification, frame header identification, function code, data bit length and check mode.

[0074] Step S203, the key field sequences are converted into multi-dimensional feature vectors, and a protocol classification model is constructed based on a gradient boosting model combined with a Bayesian optimization algorithm; the multi-dimensional feature vectors include field distribution entropy values, jump frequency parameters and weighted feature indexes.

[0075] Step S204, a similarity matrix of the multi-dimensional feature vectors and a preset protocol feature library is calculated based on the trained protocol classification model using a cosine similarity algorithm; the preset protocol feature library stores standard feature vectors of known protocol types.

[0076] Step S205, the protocol type and version number are matched according to the similarity matrix; the version number corresponds to the maximum similarity entry in the preset protocol feature library.

[0077] Step S206, the unmatched protocol feature fields are extracted as private protocol features; the private protocol features include specific manufacturer-defined check bit patterns and custom instruction codes.

[0078] Step S207, the protocol type, version number and private protocol features are written into the preset protocol feature library to obtain a protocol feature dataset.

[0079] Specifically, first, binary original message data generated by multimedia devices in the protocol interaction process is collected through RS-232 / 485, TCP / IP and other multi-communication interfaces, and preprocessed for noise reduction, field segmentation and the like, and key fields such as protocol header identification, function code, data bit length, check mode are extracted to form a structured field sequence. Secondly, the field sequence is converted into a multi-dimensional feature vector including field distribution entropy values, jump frequency parameters and weighted feature indexes, and a protocol classification model is constructed using a gradient boosting model combined with a Bayesian optimization algorithm; the cosine similarity matrix of the feature vector and the known protocol standard feature vector in the preset protocol feature library is calculated through the model, the protocol type and the corresponding maximum similarity version number are matched, and the unmatched fields are extracted as private protocol features including manufacturer-defined check bit patterns and instruction codes. Finally, the protocol type, version number and private features are written into the preset protocol feature library to generate a protocol feature dataset.

[0080] The embodiment significantly improves the parsing efficiency and accuracy of multi-vendor equipment protocols through standardized data collection and intelligent feature analysis: supports RS-232 / 485, TCP / IP and other multi-interface protocol message collection, automatically adjusts parameters through gradient boosting model and Bayesian optimization algorithm, realizes rapid identification of heterogeneous protocols such as Modbus RTU and custom TCP, avoids high cost and low efficiency of manual analysis, and meets the access requirements of "plug and play" of multimedia devices; through field weight calculation, multi-dimensional feature vector construction and cosine similarity matching, fine identification of protocol type, version and private characteristics is realized, especially for the extraction of vendor custom protocol fields, effectively solving the problem of private protocol incompatibility in traditional solutions; the identification result is written into the preset protocol feature library, supporting automatic expansion of the knowledge base under the federal learning mechanism, so that the system can continuously evolve with the access of new devices, enhance the compatibility of unknown protocols, and ensure the long-term availability and scalability of the multimedia device management system.

[0081] In one embodiment, the protocol feature data set is subjected to protocol type identification, and if the identification result is lower than the preset confidence threshold, deep feature extraction is performed to obtain protocol classification data, which can include the following steps:

[0082] Step S301, input the multi-dimensional feature vector in the protocol feature data set into the preset classification model, and output the protocol type identification result and the corresponding confidence value.

[0083] Step S302, determine whether the confidence value is lower than the preset confidence threshold; if the confidence value is lower than the confidence threshold, the multi-dimensional feature vector is subjected to secondary processing using a convolutional neural network to extract a protocol deep feature vector.

[0084] Step S303, match the protocol deep feature vector with the protocol classification template in the preset protocol feature library to obtain a protocol classification label.

[0085] Step S304, generate protocol classification data based on the protocol classification label; the protocol classification data includes protocol type identification, confidence parameter and private feature label.

[0086] Specifically, the multi-dimensional feature vector in the protocol feature data set is input into the preset classification model (the model outputs a protocol type identification result and a corresponding trust degree value. Next, it is determined whether the trust degree value is lower than a pre-set confidence threshold (for example, 0.7). If it is lower than the threshold, it indicates that the protocol type identification result based on the multi-dimensional feature vector is less reliable, and the multi-dimensional feature vector is processed again using a convolutional neural network (CNN). Through the feature extraction capability of the CNN, deep features in the protocol data are mined to extract a protocol deep feature vector. Then, the protocol deep feature vector is matched with a protocol classification template in a preset protocol feature library to obtain a more accurate protocol classification label. Finally, protocol classification data is generated based on the protocol classification label, which contains key information such as protocol type identification, confidence parameter, and private feature label.

[0087] The embodiment significantly improves the reliability and accuracy of protocol type identification through multi-stage intelligent analysis and accurate matching: the preset classification model is used to quickly process the multi-dimensional feature vector, outputting a preliminary protocol type identification result and a trust degree value; when the trust degree value is lower than the threshold, a convolutional neural network is introduced for secondary processing to mine a protocol deep feature vector. This process can capture more subtle and representative features in the protocol data, thereby improving the accuracy of identification. By matching with the classification template in the preset protocol feature library, the protocol classification label obtained is more accurate; and the protocol classification data generated finally contains rich information such as protocol type identification, confidence parameter, and private feature label. These information not only help to accurately identify the protocol type, but also enable the system to understand the reliability of the identification result and the private features that may exist in the protocol, providing comprehensive support for further protocol adaptation, security processing, and device control, thereby enhancing the adaptability and reliability of the entire multimedia device control method.

[0088] In one of the embodiments, a national standard protocol adaptation conversion engine is called based on the protocol classification data, uniform format data is converted, national secret level encryption algorithms and bidirectional identity verification rules are loaded, and secure communication data frames are generated, which can include the following steps:

[0089] Step S401, according to the source protocol identification in the protocol classification data, a protocol analysis template in the national standard protocol adaptation conversion engine is matched, and intermediate structure data is generated.

[0090] Step S402, the intermediate structure data is standardized in data format, and is reorganized according to a preset field order and type to obtain uniform format data.

[0091] Step S403, according to the data sensitivity level of the uniform format data, a national secret algorithm type is selected, and a key factor is injected to generate an encryption parameter.

[0092] Step S404, bind the encryption parameter with the identity certificate in the two-way authentication strategy, and construct the verification data packet containing the timestamp, random number and device fingerprint.

[0093] Step S405, perform asymmetric encryption operation on the verification data packet through the two-way authentication strategy, and generate the secure communication data frame carrying the digital signature.

[0094] Further, according to the source protocol identifier in the protocol classification data, the protocol analysis template (such as the GB / T32960 standard template) in the national standard protocol adaptation conversion engine is matched to generate intermediate structure data. The intermediate structure data is subjected to format standardization processing, and is reorganized according to the preset field order and type to form uniform format data. According to the data sensitivity level of the uniform format data, the corresponding national secret algorithm type (such as SM2 / SM3 / SM4) is selected, and the encryption parameter is generated by injecting the key factor. The encryption parameter is bound with the identity certificate (device certificate, user Token) in the two-way authentication strategy, and a verification data packet containing a timestamp, a random number and a device fingerprint is constructed. The verification data packet is subjected to asymmetric encryption operation through the two-way authentication strategy, and a secure communication data frame carrying a digital signature is generated.

[0095] The embodiment realizes unified adaptation and trusted transmission of heterogeneous protocols through standardized conversion and security reinforcement mechanism: based on the national standard protocol template, the conversion of heterogeneous protocols to uniform format data is realized, the differences of multiple manufacturer protocols are eliminated, different brand devices are supported to access the unified management platform, and the system compatibility and scalability are improved; through national secret algorithm encryption, two-way identity authentication and digital signature mechanism, the confidentiality, integrity and anti-repudiation of data transmission are guaranteed, the security requirements of EP 2.0 are met, and the communication security sensitive scenes such as education and government are suitable.

[0096] In one of the embodiments, the hash check is performed on the secure communication data frame, and the data stream passing the integrity verification is parsed into structured interaction information, the behavior habit features are extracted combined with the user historical operation records, and the preliminary interaction information package is generated, which can include the following steps:

[0097] Step S501, a preset hash algorithm is used to perform hash check on the secure communication data frame to generate a check result.

[0098] Step S502, according to the check result, the data integrity is judged, if the check passes, according to the preset interaction protocol template, the data stream is disassembled into operation instruction field and timestamp sequence.

[0099] Step S503, the operation frequency parameter and instruction type distribution data in the user historical operation records are extracted, and the operation interval feature value is calculated combined with the timestamp sequence.

[0100] Step S504, input the operation instruction field and the operation interval characteristic value into the hidden Markov model, output the dynamic weight parameter and the behavior mode label.

[0101] Step S505, adjust the parsing priority of the interactive protocol template according to the dynamic weight parameter, and generate an interactive information package carrying the behavior mode label.

[0102] Specifically, a preset hash algorithm (such as SHA-256) is used to perform hash check on the secure communication data frame to generate a check result to judge the data integrity. If the check is passed, the data stream is disassembled into an operation instruction field and a timestamp sequence according to a pre-set interactive protocol template. The operation frequency parameter and the instruction type distribution data in the user historical operation record are extracted, and the operation interval characteristic value (such as mean value, standard deviation) is calculated in combination with the timestamp sequence. The operation instruction field and the interval characteristic value are input into the hidden Markov model, and the dynamic weight parameter and the behavior mode label (such as "high frequency operation mode") are output. The parsing priority of the interactive protocol template is adjusted according to the dynamic weight parameter, and an interactive information package carrying the behavior mode label is generated.

[0103] The embodiment improves the reliability and intelligent level of interaction control through data verification and behavior modeling mechanism: data integrity protection and structured analysis: through hash check to ensure that the transmission data is not tampered, based on the protocol template to realize the standardized disassembly of data stream, to provide accurate instruction and time sequence data support for subsequent interaction logic; use the hidden Markov model to analyze the operation rule, generate dynamic weight and behavior mode label, realize the adaptive adjustment of the parsing priority of the interactive protocol template, improve the user operation response efficiency and the personalization of control experience.

[0104] In one of the embodiments, the interactive information package is judged according to the real-time user operation feedback and the device running state data of the unified platform, and if it does not meet the current control demand, an interactive control scheme is generated through model optimization information, which can include the following steps:

[0105] Step S601, obtain real-time user operation feedback data and device running state data; the data includes user behavior feature vector and device state index matrix.

[0106] Step S602, use a preset demand matching rule to calculate the compliance of the user behavior feature vector, and obtain the compliance label of the interactive information package.

[0107] Step S603, construct a dynamic weight matrix according to the correlation degree of the compliance label and the device state index matrix.

[0108] Step S604, when the characteristic value in the dynamic weight matrix exceeds a preset optimization trigger threshold, extract the abnormal fluctuation mode in the device state data.

[0109] Step S605, input the abnormal fluctuation mode into the pre-trained proximal policy optimization model to generate a model parameter increment set for the current control scenario.

[0110] Step S606, based on the model parameter increment set, the original control strategy is updated online and iteratively to obtain an interactive control scheme containing an adaptive adjustment coefficient.

[0111] Specifically, real-time user operation feedback data and equipment running state data are obtained, wherein the data contains a user behavior feature vector and an equipment state index matrix. A preset demand matching rule is used to calculate the compliance of the user behavior feature vector, and a compliance label (such as pass, warning, and rejection) of the interactive information package is obtained. According to the correlation degree of the compliance label and the equipment state index matrix, a dynamic weight matrix is constructed to reflect the influence degree of each index on the control demand. When the eigenvalue in the dynamic weight matrix exceeds a preset optimization trigger threshold, an abnormal fluctuation mode (such as a sudden increase in response time and an error rate exceeding the limit) in the equipment state data is extracted. The abnormal fluctuation mode is input into a pre-trained proximal policy optimization (PPO) model to generate a model parameter increment set for the current control scenario. Based on the increment set, the original control strategy is updated online and iteratively to obtain an interactive control scheme containing an adaptive adjustment coefficient.

[0112] The embodiment improves the adaptive control capability of the system for complex scenarios through real-time feedback and dynamic optimization mechanism: the compliance of user behavior is quantified through demand matching rules, and the weights are dynamically allocated according to the correlation degree of the equipment state index, so as to realize accurate quantitative analysis of the control demand, and ensure that the interactive information package meets the business rules and the equipment running state; based on the online iteration triggered by the abnormal fluctuation mode, the adaptive control parameters are generated by using the proximal policy optimization algorithm, the system resource allocation and response logic are adjusted in real time, and the rapid response capability and long-term running stability of the control strategy to equipment abnormalities are improved.

[0113] In one of the embodiments, the dynamic weight matrix can be calculated and constructed by the following formula:

[0114] ;

[0115] = ;

[0116] = ;

[0117] wherein, denotes the normalized dynamic weight matrix, denotes the dynamic weight matrix, denotes the initial weight vector, , denotes a real-time adjustment matrix, = diag( ), denotes the th real-time adjustment coefficient, denotes the historical mean of the th indicator, denotes the standard deviation of the th indicator, denotes an adjustment intensity coefficient, denotes an anomaly threshold coefficient.

[0118] Preferably, the initial weight vector is calculated based on the Pearson correlation coefficient between the device status indicators and the compliance label , reflecting the static correlation strength of the indicators and system compliance. The formula is = , that is, by normalizing the absolute value of the indicator correlation degree, the sum of the initial weights is 1, ensuring that high correlation indicators (such as error rate, response time) obtain higher weights.

[0119] The dynamic weight modeling method in this embodiment improves the system's adaptive ability to complex scenarios through data-driven and real-time adjustment: based on statistical correlation, the indicator weight is objectively determined, avoiding subjective assignment bias, so that the system pays more attention to core indicators that have a significant impact on compliance (such as error rate and compliance negatively correlated weight automatically increased); by setting an anomaly threshold based on historical data, the weight is automatically enhanced for sudden indicator fluctuations (such as a sudden increase in response time caused by device failure), guiding the system to quickly locate and handle key anomalies and shorten fault response time; normalization ensures that weight allocation does not deviate too much from overall demand due to a single indicator anomaly, and automatically returns to an equilibrium state after anomaly repair, ensuring long-term stability and reliability of the system and control strategy.

[0120] In one embodiment, as shown in Figure 2 , the application also provides a multimedia device control system based on adaptive protocol matching, which can include:

[0121] A protocol feature analysis module 701 is configured to obtain communication protocol data of the multimedia device, construct a feature vector space, and compare it with a preset protocol feature library to determine protocol type, version information, and manufacturer private protocol features, and obtain protocol feature data sets; and is further configured to perform protocol type identification on the protocol feature data sets, and if the identification result is lower than a preset confidence threshold, perform deep feature extraction to obtain protocol classification data.

[0122] A protocol security reinforcement module 702 is configured to call a national standard protocol adaptation conversion engine based on the protocol classification data, convert the protocol classification data into unified format data, load national secret level encryption algorithms and bidirectional identity verification rules, and generate secure communication data frames.

[0123] The interaction information processing module 703 is configured to perform hash check on the secure communication data frame, parse the data stream passing the integrity verification into structured interaction information, extract the behavior habit features in combination with the user historical operation records, and generate a preliminary interaction information package.

[0124] The feedback control optimization module 704 is configured to judge the interaction information package according to the real-time user operation feedback and the equipment running state data of the unified platform, and generate an interaction control scheme through model optimization information if the interaction information package does not meet the current control requirement.

[0125] The above-mentioned multimedia device control system based on adaptive protocol matching realizes adaptive control of the multimedia device through cooperation of various modules of the system: the protocol feature analysis module collects device original protocol messages through multiple communication interfaces, extracts key feature fields and converts them into a multi-dimensional feature vector, determines the protocol type, version and manufacturer private features after comparison with a preset protocol feature library, generates a protocol feature data set, and performs confidence check on the identification result, performs deep feature extraction on the low-confidence result through a convolutional neural network to obtain protocol classification data; the protocol security reinforcement module calls a national standard protocol adaptation conversion engine based on the protocol classification data, converts heterogeneous protocols into unified format data, loads national secret level encryption algorithms and bidirectional identity verification rules, and generates secure communication data frames meeting security standards; the interaction information processing module performs hash check on the secure communication data frames, parses the data stream passing the check into structured interaction information, extracts behavior habit features in combination with user historical operation records, and generates a preliminary interaction information package; the feedback control optimization module judges the compliance of the interaction information package according to real-time user operation feedback and equipment running state data, identifies abnormal fluctuations through construction of a dynamic weight matrix, updates the control strategy online using an optimization model, and generates an adaptive interaction control scheme. The system builds a complete technical link from protocol analysis to control optimization through modular design: the protocol feature analysis module cooperates with the protocol security reinforcement module to realize automatic identification, standardized conversion and security reinforcement of protocols of devices of multiple manufacturers, solves the problems of poor protocol compatibility and high manual adaptation cost in traditional schemes, and meets the security requirements of the Cybersecurity Protection 2.0; the interaction information processing module cooperates with the feedback control optimization module to dynamically adjust the interaction logic and control strategy through user behavior modeling and real-time state feedback, improves the intelligent level of device response and the self-adaptive ability to complex scenes, and ensures the efficiency, reliability and security of multimedia device control.

[0126] It should be understood that, although the steps in the flowcharts related to the embodiments described above are shown in sequence according to the arrows, the steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the execution of the steps is not strictly limited in sequence, and the steps can be executed in other sequences. Moreover, at least some of the steps in the flowcharts related to the embodiments described above can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of the steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least some of the other steps or the steps or stages in the other steps.

[0127] In an embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor implementing the steps of the adaptive protocol matching based multimedia device control method and system as described above when executing the computer program.

[0128] In an embodiment, a computer readable storage medium is provided, storing a computer program, and the computer program implementing the steps of the method embodiments described above when executed by a processor.

[0129] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts are described with reference to the parts of the method embodiments. The device embodiments described above are merely illustrative, and the components described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e. they can be located in one place or distributed on multiple network units. Some or all of the modules can be selected to achieve the purposes of the present disclosure according to actual needs. Those skilled in the art can understand and implement it without creative labor.

[0130] The above-described embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the patent scope of the application. It should be pointed out that, for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the protection scope of the present application.

Claims

1. A method for controlling a multimedia device based on adaptive protocol matching, the method comprising: The method comprises: Obtaining communication protocol data of a multimedia device, constructing a feature vector space and comparing with a preset protocol feature library, determining protocol type, version information and manufacturer private protocol features, and obtaining protocol feature data set; Performing protocol type identification on the protocol feature data set, and if the identification result is lower than a preset confidence threshold, performing deep feature extraction to obtain protocol classification data; Based on the protocol classification data, calling a national standard protocol adaptation conversion engine, converting into unified format data, loading national secret level encryption algorithm and bidirectional identity verification rules, and generating secure communication data frame; Performing hash check on the secure communication data frame, and parsing data stream passing integrity verification into structured interactive information, combining user historical operation record to extract behavior habit features, and generating preliminary interactive information package; According to real-time user operation feedback and device running state data of the unified platform, judging the interactive information package, and if it does not meet the current control demand, generating interactive control scheme through model optimization information; The method comprises: Obtaining real-time user operation feedback data and device running state data; the data contains user behavior feature vector and device state index matrix; performing compliance calculation on the user behavior feature vector by using preset demand matching rules to obtain compliance label of the interactive information package; and constructing dynamic weight matrix according to the correlation degree of the compliance label and the device state index matrix; When the feature value in the dynamic weight matrix exceeds a preset optimization trigger threshold, extract the abnormal fluctuation mode in the device running state data; input the abnormal fluctuation mode into a pre-trained proximal strategy optimization model to generate a model parameter increment set for the current control scene; based on the model parameter increment set, perform online iterative update on the original control strategy to obtain an interactive control scheme containing an adaptive adjustment coefficient.

2. The method of claim 1, wherein, The method comprises: Obtaining multimedia device communication protocol original message data; the original message data contains binary stream in protocol interaction process; Pretreating the original message data, and extracting key field sequence according to the pretreated protocol message; the key field sequence contains protocol header identifier, frame header identifier, function code, data bit length and check mode; Convert the key field sequence into a multi-dimensional feature vector, and construct a protocol classification model based on gradient boosting model combined with Bayesian optimization algorithm; the multi-dimensional feature vector contains field distribution entropy value, jump frequency parameter and weighted feature index; Based on the trained protocol classification model, calculate the similarity matrix of the multi-dimensional feature vector and the preset protocol feature library by using cosine similarity algorithm; the preset protocol feature library stores standard feature vector of known protocol type; According to the similarity matrix matching protocol type and version number; the version number corresponds to the maximum similarity entry in the preset protocol feature library; Extracting the unmatched protocol feature field as a private protocol feature; the private protocol feature includes a specific manufacturer-defined check bit pattern and a self-defined instruction code; Write the protocol type, version number and private protocol feature into the preset protocol feature library to obtain a protocol feature dataset.

3. The method of claim 1, wherein, The protocol type recognition is performed on the protocol feature dataset, and if the recognition result is lower than a preset confidence threshold, deep feature extraction is performed to obtain protocol classification data, including: Inputting a multi-dimensional feature vector in the protocol feature dataset into a preset classification model outputs a protocol type recognition result and a corresponding confidence value; Judging whether the confidence value is lower than a preset confidence threshold; if the confidence value is lower than the confidence threshold, performing secondary processing on the multi-dimensional feature vector using a convolutional neural network to extract a protocol deep feature vector; According to the protocol deep feature vector, matching the protocol classification template in the preset protocol feature library to obtain a protocol classification label; Generating protocol classification data based on the protocol classification label; the protocol classification data includes a protocol type identifier, a confidence parameter and a private feature marker.

4. The method of claim 1, wherein, The protocol classification data is used to call a national standard protocol adaptation conversion engine, convert into unified format data, load a national secret level encryption algorithm and a two-way identity verification rule, generate a secure communication data frame, including: According to the source protocol identifier in the protocol classification data, matching the protocol analysis template in the national standard protocol adaptation conversion engine to generate intermediate structure data; Standardizing the data format of the intermediate structure data and reorganizing according to a preset field order and type to obtain unified format data; According to the data sensitivity level of the unified format data, selecting a national secret algorithm type, injecting a key factor to generate an encryption parameter; Binding the encryption parameter and the identity certificate in the two-way verification strategy to construct a verification data packet containing a timestamp, a random number and a device fingerprint; Performing asymmetric encryption operation on the verification data packet through the two-way verification strategy to generate a secure communication data frame carrying a digital signature.

5. The method of claim 1, wherein, Performing hash check on the secure communication data frame, and parsing the data stream that passes the integrity verification into structured interaction information, extracting behavior habit features combined with user historical operation records to generate a preliminary interaction information packet, including: Performing hash check on the secure communication data frame using a preset hash algorithm to generate a check result; According to the check result, judging the data integrity, if the check passes, according to a preset interaction protocol template, disassembling the data stream into an operation instruction field and a timestamp sequence; Extracting the operation frequency parameter and instruction type distribution data in the user historical operation records, and calculating the operation interval feature value combined with the timestamp sequence; Inputting the operation instruction field and the operation interval feature value into a hidden Markov model to output dynamic weight parameters and behavior mode labels; According to the dynamic weight parameters, adjusting the parsing priority of the interaction protocol template to generate an interaction information packet carrying a behavior mode label.

6. The method of claim 1, wherein, The dynamic weight matrix is constructed by the following formula: ； = ； = ； wherein, denotes the normalized dynamic weight matrix, denotes the dynamic weight matrix, denotes the initial weight vector, , denotes the real-time adjustment matrix, = diag( ), denotes the th real-time adjustment coefficient, denotes the historical mean of the th index, denotes the standard deviation of the th index, denotes the adjustment intensity coefficient, denotes the abnormal threshold coefficient.

7. A multimedia device control system based on adaptive protocol matching, characterized by, The system comprises: The protocol feature analysis module is configured to acquire communication protocol data of the multimedia device, construct a feature vector space, compare with a preset protocol feature library, determine protocol types, version information, and manufacturer private protocol features, and obtain a protocol feature dataset; and is further configured to perform protocol type identification on the protocol feature dataset, and if the identification result is lower than a preset confidence threshold, perform deep feature extraction to obtain protocol classification data. The protocol security reinforcement module is configured to call a national standard protocol adaptation conversion engine based on the protocol classification data, convert the protocol classification data into unified format data, load a national secret level encryption algorithm and a two-way identity verification rule, and generate a secure communication data frame. The interactive information processing module is configured to perform hash check on the secure communication data frame, parse a data stream that passes integrity verification into structured interactive information, extract behavior habit features in combination with user historical operation records, and generate a preliminary interactive information package. The feedback control optimization module is configured to judge the interactive information package according to real-time user operation feedback and device running state data of the unified platform, and if the interactive information package does not meet current control requirements, generate an interactive control scheme through model optimization information. The feedback control optimization module is configured to judge the interactive information package according to real-time user operation feedback and device running state data of the unified platform, and if the interactive information package does not meet current control requirements, generate an interactive control scheme through model optimization information. The feedback control optimization module is configured to judge the interactive information package according to real-time user operation feedback and device running state data of the unified platform, and if the interactive information package does not meet current control requirements, generate an interactive control scheme through model optimization information. The feedback control optimization module is configured to judge the interactive information package according to real-time user operation feedback and device running state data of the unified platform, and if the interactive information package does not meet current control requirements, generate an interactive control scheme through model optimization information.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes the computer program to realize the steps of the method of any one of claims 1 to 6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the method of any one of claims 1 to 6.

Images