Brake drum heat treatment process method based on digital twin drive

Abstract

The invention discloses a brake drum heat treatment process method based on digital twin drive, which comprises the following steps: acquiring real-time process parameters including temperature, atmosphere and heating time from a heat treatment furnace group, acquiring and transmitting the real-time process parameters to a central data processing module through a sensor network to obtain a standardized parameter data set; according to the synchronous adjustment instruction, the temperature curve and atmosphere control parameters of each heat treatment furnace are updated in real time, cooperative adjustment is achieved through a communication protocol between devices, and a consistent process execution state is obtained; extracting sensor data from the consistent process execution state, and generating a data record binding a physical entity and a digital model in combination with a timestamp and a device identifier to obtain a traceable data set; and aiming at the traceable data set, encrypting and storing the mapping relation between the sensor data and the process result by adopting a block chain technology, judging whether the data is tampered or not, and obtaining a tamper-proof verification result.

Description

Technical Field

[0001] This invention belongs to the field of information technology, specifically a heat treatment process for brake drums based on digital twin drive. Background Technology

[0002] In the modern automotive manufacturing industry, brake drums are a core component ensuring driving safety, and their quality stability directly affects vehicle operation safety. With the continuous increase in automobile production and increasingly stringent quality standards, the collaborative production model of heat treatment furnace clusters has become the mainstream choice for achieving large-scale, efficient production of brake drums. This model, through the parallel operation of multiple heat treatment furnaces, can significantly improve production efficiency, but it also places extremely high demands on the process coordination and quality control among the furnace clusters. It requires ensuring precise synchronization and real-time optimization of process parameters for each furnace in a dynamic and complex production environment, and achieving reliable traceability of data throughout the entire process, thereby guaranteeing a high degree of consistency in the performance and quality indicators of brake drum materials.

[0003] However, the current collaborative production technology system for heat treatment furnace groups still faces multiple challenges. First, at the level of process parameter synchronization, due to the distributed operation of the furnace group, key parameters such as temperature, atmosphere, and heating time of each furnace are easily affected by factors such as fluctuations in sensor accuracy, differences in equipment aging, and interference from the workshop environment (such as unstable voltage, temperature and humidity changes), making it difficult to maintain real-time consistency. For example, a furnace may experience a deviation from the standard process window in its actual temperature curve due to thermocouple response delay or heating tube power attenuation, resulting in uneven grain size or hardness values ​​exceeding tolerances in the metallographic structure of the brake drum, directly compromising the uniformity of product quality.

[0004] Secondly, although distributed consensus algorithms are used to coordinate parameter adjustments across multiple furnaces, they face bottlenecks in balancing computational efficiency and real-time performance. Excessive iteration cycles or instruction transmission delays can cause parameter deviations in some furnaces to exceed permissible thresholds before adjustment, weakening the collaborative control effect. Furthermore, in the data trust management phase, the generation of the entire process dataset requires precise binding of real-time sensor data with timestamps and unique equipment identifiers. However, there is a contradiction between storing massive amounts of data and blockchain encryption verification; excessive pursuit of data tamper-proofing can lead to a surge in storage costs, and the time spent on encryption and decryption can affect the response speed of real-time quality analysis. If data is tampered with during acquisition, transmission, or storage, or if the identifier mapping relationship is incorrect, it will directly distort the correlation analysis between historical process parameters and product quality indicators, causing machine learning-based regression optimization models to output incorrect strategies, creating a vicious cycle of "data pollution, model failure, and process out of control."

[0005] Finally, during the transmission and execution of dynamically optimized parameter commands between devices, issues such as industrial Ethernet communication latency and insufficient protocol compatibility between different brands of furnaces can affect the process. This leads to a discrepancy between the actual adjustments of the physical equipment and the expected state of the digital twin model, reducing the guiding value of virtual simulation for actual production. These problems essentially all point to the same core issue: how to build a technical system with both precise collaborative capabilities and reliable management mechanisms in a distributed, multi-variable, and high-real-time production scenario. This is precisely the key bottleneck that urgently needs to be overcome in the current collaborative production braking drum process of heat treatment furnace groups. Summary of the Invention

[0006] The purpose of this invention is to provide a brake drum heat treatment process based on digital twin drive, so as to solve the problems in the prior art mentioned in the background, such as the difficulty in synchronizing process parameters due to the distributed operation characteristics, the difficulty in balancing the real-time performance and efficiency of the algorithm, the difficulty in balancing massive data storage and encryption verification, and the susceptibility of optimized instruction transmission and execution to interference.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A heat treatment process for brake drums based on digital twin drive includes: Step S1: Obtain real-time process parameters from the heat treatment furnace group, collect them through the sensor network and transmit them to the central data processing module to obtain a standardized parameter dataset; Step S2: For the standardized parameter dataset collected in step S1, a distributed consensus algorithm is used to coordinate the parameter synchronization of multiple heat treatment furnaces, and it is determined whether the parameter deviation of each device exceeds the preset threshold to obtain a synchronization adjustment instruction. Step S3: Based on the synchronization adjustment instructions obtained in step S2, update the temperature curve and atmosphere control parameters of each heat treatment furnace in real time, and achieve coordinated adjustment through the inter-equipment communication protocol to obtain a consistent process execution status. Step S4: Extract sensor data from the consistent process execution status, combine it with timestamps and equipment identifiers to generate data records that bind physical entities and digital models, and obtain a traceable dataset. Step S5: For the traceable dataset, use blockchain technology to encrypt and store the mapping relationship between sensor data and process results, determine whether the data has been tampered with, and obtain the tamper-proof verification result. Step S6: Based on the anti-tampering verification results, extract historical process parameters and brake drum quality indicators from the traceable dataset, and use machine learning regression algorithm to analyze the correlation between parameters and quality to obtain an optimized parameter adjustment strategy. Step S7: Based on the optimized parameter adjustment strategy, update the control parameters of the heat treatment furnace group in real time, distribute the strategy through the inter-equipment collaborative communication network, determine whether the adjusted parameters meet the production consistency requirements, and obtain the dynamically optimized process execution status. Step S8: Extract real-time sensor data from the dynamically optimized process execution status, update the parameter mapping of the digital model, determine whether the synchronization deviation between the model and the physical entity is within the preset range, and obtain the synchronization verification result. Step S9: Based on the synchronization verification results, adjust the feedback control logic of the central data processing module, generate new process parameter synchronization instructions, and transmit them to the heat treatment furnace group to obtain the updated standardized parameter dataset.

[0008] According to the above technical solution, in step S1, data is collected and transmitted to the central data processing module through a sensor network to obtain a standardized parameter dataset, including: Step S101: Real-time process parameters are obtained from the heat treatment furnace group through the sensor network. The process parameters include temperature data, atmosphere data and heating time. The data are transmitted to the central data processing module to obtain the raw parameter dataset. Step S102: The original parameter dataset is preprocessed using a data cleaning algorithm to remove outliers and missing values, resulting in a cleaned parameter dataset. If the cleaned parameter dataset meets the preset integrity threshold, the temperature data, atmosphere data, and heating time are normalized through standardization to obtain a standardized parameter dataset. Step S103: Based on the standardized parameter dataset, the K-means clustering algorithm is used to classify the furnace group status to obtain a set of operating modes for process parameters; if a mode in the set of operating modes deviates from the preset process specification threshold, the parameter adjustment trend is predicted by regression analysis algorithm to obtain a set of adjustment parameters. Step S104: Based on the set of adjustment parameters, generate process parameter optimization instructions for the furnace group, transmit them to the heat treatment furnace group control system, and obtain the optimized furnace group operating status. Step S105: By monitoring the optimized furnace group's operating status in real time, new process parameters are collected and the standardized parameter dataset is updated cyclically.

[0009] According to the above technical solution, in step S2, for the standardized parameter dataset, a distributed consensus algorithm is used to coordinate the parameter synchronization of multiple heat treatment furnaces, determine whether the parameter deviation of each device exceeds a preset threshold, and obtain a synchronization adjustment instruction, including: Step S201: Obtain the operating data of multiple heat treatment furnaces and generate a standardized dataset through standardization processing; Step S202: Using a distributed consensus algorithm, extract the parameters of each device from the standardized dataset and perform parameter synchronization operation; if any device parameter is found to be inconsistent with the master node parameter during the parameter synchronization process, calculate the parameter deviation; and compare the parameter deviation with a preset threshold to determine whether the deviation exceeds the threshold range. Step S203: Based on the deviation judgment result, generate a synchronization instruction, which includes specific adjustment instructions; Step S204: Using a distributed consensus algorithm, the adjustment instructions are distributed to the corresponding heat treatment furnaces to perform parameter adjustments; Step S205: Collect new parameters from the adjusted heat treatment furnace, verify the execution effect of the synchronization command, and generate verification results.

[0010] According to the above technical solution, in step S3, the temperature curve and atmosphere control parameters of each heat treatment furnace are updated in real time according to the synchronization adjustment command. Coordinated adjustment is achieved through inter-equipment communication protocols to obtain a consistent process execution status, including: Step S301: Obtain the real-time temperature curves and atmosphere control parameters of each heat treatment furnace through the inter-device communication protocol to obtain a unified dataset; If the temperature curve deviation in the dataset exceeds the preset threshold, the heating power of each furnace is corrected through a synchronous adjustment algorithm to determine a consistent temperature curve. Step S302: Based on the corrected temperature curve, adjust the oxygen content and pressure parameters of each furnace using a preset atmosphere control model to obtain optimized atmosphere control parameters. Step S303: The optimized temperature curve and atmosphere control parameters are transmitted to each heat treatment furnace through the inter-equipment communication protocol to obtain control commands for coordinated execution. Step S304: If there is a deviation between the control command executed in coordination and the current state inside the furnace, the command is fine-tuned through a feedback adjustment algorithm to determine a consistent process execution state. Step S305: Obtain the fine-tuned control command and distribute it to each heat treatment furnace through the inter-equipment communication protocol to obtain the final process execution status.

[0011] According to the above technical solution, in step S4, sensor data is extracted from the consistent process execution status, and combined with timestamps and equipment identifiers to generate data records that bind physical entities and digital models, resulting in a traceable dataset, including: Step S401: Obtain real-time sensor data from the sensor, fuse timestamp records and device identifiers to generate raw data records containing unique identifiers; If the sensor data is missing timestamp records or device identifiers, invalid data will be discarded according to the preset verification rules to obtain complete original data records; Step S402: The original data records are processed using a data fusion algorithm, and the sensor data are time-series aligned based on the timestamp records to generate standardized data records. Step S403: Through preset mapping rules, standardized data records are associated with physical entities and bound to the corresponding digital models to obtain the data records bound to the digital models; If the data records bound to the digital model are inconsistent with the physical entity state, the support vector machine algorithm is used to detect anomalies in the data records and identify abnormal data records. Step S404: Based on the anomaly detection results, update the state parameters of the digital model and generate traceable data records reflecting the process status. Step S405: Index the traceable data records using timestamp records and device identifiers to obtain a traceable dataset that supports fast querying.

[0012] According to the above technical solution, in step S5, for the traceable dataset, blockchain technology is used to encrypt and store the mapping relationship between sensor data and process results, determine whether the data has been tampered with, and obtain the tamper-proof verification result, including: Step S501: Collect data through sensors, generate a mapping relationship between sensor data and process results, calculate the hash value of the mapping relationship using a hash algorithm, and obtain the initial hash record; Step S502: Store the initial hash record to the blockchain, generate on-chain records using distributed ledger technology, and obtain the hash value stored on the blockchain; Step S503: Obtain on-chain records from the blockchain, extract the stored hash values, and obtain on-chain hash data; Step S504: Collect data again through the sensor, generate a new mapping relationship, calculate the hash value of the new mapping relationship using the same hash algorithm, and obtain the current hash value; If the on-chain hash data matches the current hash value, it is determined that the data has not been tampered with, and the data integrity verification result is obtained; If the on-chain hash data is inconsistent with the current hash value, the generation time of the on-chain record and the current data are compared and analyzed using timestamps to determine the time point when the tampering occurred and obtain the tampering verification result. Step S505: Based on the tamper verification result, use digital signature technology to encrypt and store the untampered data, generate a new on-chain record, and obtain the tamper-proof storage result.

[0013] According to the above technical solution, in step S6, based on the anti-tampering verification results, historical process parameters and brake drum quality indicators are extracted from the traceable dataset. A machine learning regression algorithm is then used to analyze the correlation between the parameters and quality, resulting in an optimized parameter adjustment strategy, including: Step S601: Extract historical process parameters and brake drum quality indicators from the traceable dataset, construct a structured dataset containing parameters and quality, and obtain the initial dataset; Step S602: The initial dataset is preprocessed using data cleaning techniques to remove missing values ​​and outliers, resulting in a cleaned dataset. Step S603: Train the cleaned dataset using a linear regression algorithm, analyze the correlation between process parameters and brake drum quality, and obtain a regression model; If the prediction error of the regression model is lower than the preset threshold, then the process parameters that significantly affect quality are extracted from the regression model to obtain the set of key parameters. If the prediction error exceeds the preset threshold, the random forest algorithm is used to retrain the model to obtain an updated regression model and set of key parameters. Step S604: Based on the set of key parameters, calculate the weighted influence of each parameter on the quality of the brake drum to obtain the parameter weight distribution; Step S605: Through parameter weight distribution, the gradient descent method is used to optimize the process parameter values ​​to obtain the optimized parameter adjustment strategy; Step S606: Based on the optimized parameter adjustment strategy, generate a configuration file for adjusting process parameters and output the adjusted parameter values.

[0014] According to the above technical solution, in step S7, the control parameters of the heat treatment furnace group are updated in real time according to the optimized parameter adjustment strategy. The adjusted parameters are then distributed through an inter-equipment collaborative communication network to determine whether they meet production consistency requirements, thereby obtaining the dynamically optimized process execution status, including: Step S701: Obtain real-time process parameters from the heat treatment furnace group through the collaborative communication network, store them in the pre-established process database, and obtain a set of real-time parameters; Step S702: Based on the real-time parameter set, the support vector machine algorithm is used to analyze the parameter fluctuation characteristics and determine the changing trend of key process parameters; If the trend of change exceeds the preset threshold, a new parameter configuration is generated through dynamic adjustment strategy to obtain an optimized parameter set; Step S703: Distribute the optimized parameter set to each device in the heat treatment furnace group through the parameter distribution mechanism, update the device control parameters, and obtain the updated parameter execution status; Step S704: Obtain the updated parameter execution status, use logistic regression algorithm to determine whether the parameters meet the production consistency requirements, and obtain the consistency evaluation result; Step S705: Based on the consistency assessment results, if the consistency is not met, adjust the transmission priority of the cooperative communication network, optimize the parameter distribution mechanism, and obtain the improved distribution strategy. Step S706: Redistribute the set of optimized parameters using the improved distribution strategy, obtain the new process execution state, and determine whether the dynamic optimization state is stable.

[0015] According to the above technical solution, in step S8, real-time sensor data is extracted from the dynamically optimized process execution state, the parameter mapping of the digital model is updated, and it is determined whether the synchronization deviation between the model and the physical entity is within a preset range to obtain the synchronization verification result, including: Step S801: Obtain real-time sensor data from the process execution status, and extract temperature, pressure and flow parameters at a preset frequency through the data acquisition module to obtain real-time sensor data; Step S802: Based on real-time sensor data, update the digital model parameters using a parameter mapping algorithm, and adjust the model parameters using a linear regression method to obtain the updated digital model parameters; Step S803: Compare the updated digital model parameters with the physical entity state, and calculate the synchronization deviation between them using the Euclidean distance formula to obtain the synchronization deviation value. The Euclidean distance formula is as follows: In the formula, Indicates the parameter values ​​of the digital model. Represents the state value of a physical entity; If the synchronization deviation value is less than the preset range threshold, the synchronization status is determined to be normal, and the synchronization verification result is obtained. If the synchronization deviation value is greater than or equal to the preset range threshold, the parameter adjustment mechanism is triggered, and a synchronization abnormality signal is obtained. Step S804: Based on the synchronization anomaly signal, the gradient descent algorithm is used to optimize the digital model parameters. The synchronization deviation is reduced through iterative calculation to obtain the optimized digital model parameters. Step S805: Using the optimized digital model parameters, recalculate the synchronization deviation with the physical entity state, and obtain a new synchronization deviation value using the Euclidean distance formula. If the new synchronization deviation value is less than the preset range threshold, the synchronization verification result is output. If the new synchronization deviation value is still greater than or equal to the preset range threshold, the deviation analysis data is recorded to obtain the final verification result.

[0016] According to the above technical solution, in step S9, based on the synchronization verification results, the feedback control logic of the central data processing module is adjusted to generate new process parameter synchronization instructions, which are then transmitted to the heat treatment furnace group to obtain an updated standardized parameter dataset, including: Step S901: Obtain deviation data from the synchronous verification results, use statistical analysis methods to determine the range of deviation values, and obtain the basis for parameter adjustment; Step S902: Adjust the feedback control logic of the central data processing module according to the deviation range, update the control parameters using the PID algorithm, and obtain the optimized control logic; Step S903: Generate process parameter synchronization instructions from the optimized control logic, format the instruction content using a data encapsulation protocol, and obtain a standardized instruction set; Step S904: Transmit the standardized instruction set to the heat treatment furnace group through the parameter synchronization transmission protocol. If data loss is detected during transmission, the instruction is resent to obtain a confirmed transmission status. Step S905: Obtain real-time feedback data from the heat treatment furnace group, use a data verification algorithm to determine the integrity of the feedback data, and obtain the verified feedback dataset. Step S906: Update the standardized parameter dataset based on the verified feedback dataset. If the parameter deviation exceeds the preset threshold, re-trigger the control logic adjustment to obtain the final parameter dataset. Step S907: By comparing the final parameter dataset with the initial synchronization verification results, the consistency analysis method is used to judge the parameter synchronization effect and obtain the optimization results of the standardized process parameters.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This invention addresses the logical correlation between parameter synchronization, data traceability, and process optimization across multiple heat treatment furnaces. It constructs a standardized dataset by real-time acquisition of process parameters such as temperature and atmosphere, employs a distributed consensus algorithm to coordinate equipment parameter synchronization, and generates adjustment instructions to ensure production consistency. Simultaneously, it utilizes blockchain to encrypt and store sensor data mapped to process results, ensuring data tamper-proofing. Furthermore, it uses machine learning regression analysis to analyze the correlation between historical parameters and brake drum quality, generating optimization adjustment strategies. These strategies are distributed through a collaborative communication network, updating furnace group parameters in real-time, dynamically optimizing process execution status, and adjusting feedback control logic through synchronous verification between digital models and physical entities, generating new synchronization instructions.

[0018] This invention achieves efficient synchronization of process parameters, secure and traceable data, and unified quality optimization, significantly improving the stability of the heat treatment process and product quality. Attached Figure Description

[0019] Figure 1 This is a flow chart of the heat treatment process of the present invention; Figure 2 Flowchart for obtaining the standardized parameter dataset for this invention; Figure 3 A flowchart for obtaining synchronization adjustment instructions for this invention; Figure 4 This invention obtains a flowchart of the process execution status to achieve consistency. Figure 5 Flowchart for obtaining traceable datasets in this invention; Figure 6 This is a flowchart illustrating the process of obtaining tamper-proof verification results for this invention. Figure 7 This invention provides a flowchart for obtaining the optimization parameter adjustment strategy. Figure 8 To obtain a dynamically optimized process execution status flowchart for this invention; Figure 9 This is a flowchart illustrating the process of obtaining synchronous verification results for this invention. Figure 10 The flowchart for obtaining the updated standardized parameter dataset for this invention is shown below. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1 like Figures 1 to 10 As shown, a specific heat treatment process for a brake drum based on digital twin drive in this embodiment may include: Step 1: Obtain real-time process parameters from the heat treatment furnace group, including temperature, atmosphere, and heating time. Collect these parameters through a sensor network and transmit them to the central data processing module to obtain a standardized parameter dataset.

[0022] like Figure 2As shown, real-time process parameters, including temperature data, atmosphere data, and heating time, are acquired from the heat treatment furnace group via a sensor network and transmitted to the central data processing module to obtain the raw parameter dataset. A data cleaning algorithm is used to preprocess the raw parameter dataset, removing outliers and missing values ​​to obtain a cleaned parameter dataset. If the cleaned parameter dataset meets a preset integrity threshold, the temperature data, atmosphere data, and heating time are normalized through standardization to obtain a standardized parameter dataset. Based on the standardized parameter dataset, the furnace group status is classified using a K-means clustering algorithm to obtain a set of operating modes for the process parameters. If a mode in the operating mode set deviates from a preset process specification threshold, a regression analysis algorithm is used to predict the parameter adjustment trend, resulting in an adjusted parameter set. Based on the adjusted parameter set, process parameter optimization instructions for the furnace group are generated and transmitted to the heat treatment furnace group control system to obtain the optimized furnace group operating status. By monitoring the optimized furnace group operating status in real time, new process parameters are collected, and the standardized parameter dataset is updated cyclically.

[0023] Specifically, when obtaining real-time process parameters from the heat treatment furnace group, data acquisition can be achieved through a distributed sensor network. Each heat treatment furnace is equipped with multi-point thermocouple sensors (such as K-type thermocouples, with an accuracy of ±1.5°C) to collect temperature data at a frequency of once per second, for example, the furnace temperature distribution is 1200°C, 1180°C, and 1190°C; atmosphere sensors (such as oxygen sensors, detecting oxygen content in the range of 0-1000ppm) collect data every 2 seconds, recording an oxygen content of 50ppm; heating time is recorded using timestamps, accurate to milliseconds, for example, if the heating start time is 2025-07-24 01:00:00.000, and the current time is 01:09:00.000, the calculated heating duration is 540 seconds. Sensor data is transmitted to the central data processing module via industrial Ethernet (such as Modbus TCP protocol) at a transmission rate of 100Mbps to ensure data real-time performance. The central module employs a data standardization algorithm. First, it detects outliers in the temperature data using the Z-score method (Z=(x-μ) / σ, where μ is the mean and σ is the standard deviation). For example, if the mean temperature data is 1190°C and the standard deviation is 10°C, and 1220°C is detected as an outlier (Z=3), it is removed. Atmosphere data is padded with missing values ​​using linear interpolation; for example, the oxygen content sequence [50,48,null,47] is interpolated to [50,48,47.5,47]. Heating time data is standardized to seconds using time difference calculation. After standardization, a unified format dataset is generated, such as a JSON structure: {"temperature":1190,"oxygen":47.5,"heating_time":540}. Data analysis uses a sliding window algorithm (window size 60 seconds) to calculate the temperature fluctuation rate (standard deviation / mean = 0.0084). If the fluctuation rate is <0.01, the process is considered stable; otherwise, an alarm is triggered. All processing is automated by a central module that interfaces with the production scheduling system. If the temperature fluctuation exceeds the standard, the heating power is automatically adjusted (e.g., reduced by 10% to 900kW) to ensure stable process parameters and thus optimize the heat treatment quality.

[0024] Step 2: For the standardized parameter dataset, a distributed consensus algorithm is used to coordinate the parameter synchronization of multiple heat treatment furnaces, determine whether the parameter deviation of each device exceeds the preset threshold, and obtain the synchronization adjustment instruction.

[0025] like Figure 3As shown, operational data from multiple heat treatment furnaces is acquired and standardized to generate a standardized dataset. A distributed consensus algorithm is used to extract parameters for each device from the standardized dataset and perform parameter synchronization. If any device parameter is found to be inconsistent with the master node parameter during synchronization, the parameter deviation is calculated. The parameter deviation is compared with a preset threshold to determine if it exceeds the threshold range. Based on the deviation determination result, a synchronization command containing specific adjustment instructions is generated. The adjustment instructions are distributed to the corresponding heat treatment furnaces using the distributed consensus algorithm, and parameter adjustments are executed. New parameters are collected from the adjusted heat treatment furnaces to verify the effectiveness of the synchronization command execution and generate verification results.

[0026] Specifically, for the standardized parameter dataset, the parameters of multiple heat treatment furnaces are first collected and standardized. Assuming each furnace collects three parameters—temperature, pressure, and heating time—the dataset format is {furnace ID, temperature (°C), pressure (Pa), heating time (s)}, for example, {furnace 1, 950.5, 101325.0, 3600}. Through data preprocessing, the parameters are normalized to the [0,1] interval using a min-max normalization algorithm, with the formula X_norm=(X-X_min) / (X_max-X_min). Taking temperature as an example, assuming the temperature range is [900,1000], then 950.5 is normalized to (950.5-900) / (1000-900)=0.505. Next, a distributed consensus algorithm (such as Raft) is used to coordinate parameter synchronization. Raft ensures data consistency by electing a leader node. The leader node broadcasts a heartbeat signal every second, containing the current parameter set. Follower nodes update their local parameters upon receiving the heartbeat. If the network latency is less than 50ms, synchronization is considered successful. Parameter deviation is judged by comparing the deviation of each furnace's normalized parameters with that of the leader node. A preset threshold of 0.05 is used; if |0.505-0.510|>0.05, the deviation exceeds the limit. The analysis process is as follows: all furnace parameters are traversed, the deviation is calculated, and a synchronization adjustment command is generated. For example, if the temperature deviation of furnace 1 is 0.06, the command {Furnace 1, adjust the temperature to 951.0℃} is generated and sent to the corresponding furnace controller through the leader node. The controller adjusts the parameters based on a PID algorithm, with a proportional coefficient Kp=0.8 and an integral time Ti=100s, ensuring that the deviation converges within the threshold. After the synchronization command is issued, the execution result is verified through a log replication mechanism. If 90% of the nodes confirm that the adjustment is completed, synchronization is successful. This process is achieved through an automated system, relying on sensor data acquisition, a distributed computing framework, and control algorithms to ensure real-time consistency of parameters.

[0027] Step 3: According to the synchronization adjustment instructions, update the temperature curve and atmosphere control parameters of each heat treatment furnace in real time, and achieve coordinated adjustment through the inter-equipment communication protocol to obtain a consistent process execution status.

[0028] like Figure 4 As shown, real-time temperature curves and atmosphere control parameters of each heat treatment furnace are obtained through an inter-equipment communication protocol, resulting in a unified dataset. If the temperature curve deviation in the dataset exceeds a preset threshold, the heating power of each furnace is corrected using a synchronous adjustment algorithm to determine a consistent temperature curve. Based on the corrected temperature curve, the oxygen content and pressure parameters of each furnace are adjusted using a preset atmosphere control model to obtain optimized atmosphere control parameters. The optimized temperature curve and atmosphere control parameters are transmitted to each heat treatment furnace through the inter-equipment communication protocol to obtain coordinated control commands. If there is a deviation between the coordinated control commands and the current furnace state, the commands are fine-tuned using a feedback adjustment algorithm to determine a consistent process execution state. The fine-tuned control commands are obtained and distributed to each heat treatment furnace through the inter-equipment communication protocol to obtain the final process execution state.

[0029] Specifically, based on synchronous adjustment commands, the temperature and atmosphere data of each heat treatment furnace are acquired in real time through an industrial IoT platform. Assuming there are three heat treatment furnaces, the target temperature curve is: heating to 800°C in 0-30 minutes, holding at 800°C for 30-60 minutes, and cooling to 400°C in 60-90 minutes. The atmosphere control parameters are: nitrogen flow rate 10L / min, oxygen content ≤0.5%. First, the real-time temperature of each furnace is collected, and thermocouple data is read from the PLC via the Modbus TCP protocol. Assuming furnace 1 temperature is 795°C, furnace 2 is 805°C, and furnace 3 is 790°C, the deviations from the target temperature are calculated as -5°C, +5°C, and -10°C, respectively. A PID algorithm is used to adjust the heating power, setting the proportional coefficient Kp=0.8, integral time Ti=100s, and derivative time Td=20s. The calculation shows that the heating power of furnace 1 increases by 4%, furnace 2 decreases by 3%, and furnace 3 increases by 6%. Simultaneously, atmospheric parameters were monitored. Assuming the oxygen content in furnace 1 was 0.6%, the nitrogen flow rate was automatically adjusted to 11 L / min via a proportional valve, reducing the oxygen content to 0.4%. The furnaces communicated with each other via the OPC UA protocol, sharing the adjusted parameters. The slopes of the temperature curves of the three furnaces were compared: furnace 1 had a heating slope of 26.5°C / min, furnace 2 26.7°C / min, and furnace 3 26.3°C / min. A deviation of less than ±0.5°C / min was considered consistent. If the deviation exceeded the limit, the main control system was triggered to reallocate power and iteratively adjust until the consistency requirements were met. Finally, data analysis confirmed that the temperature fluctuation of the three furnaces during the 60-minute holding period was ≤±2°C, the atmospheric parameters were stable, and the process execution consistency reached over 98%.

[0030] Step 4: Extract sensor data from the consistent process execution status, combine it with timestamps and equipment identifiers to generate data records that bind physical entities and digital models, and obtain a traceable dataset.

[0031] like Figure 5 As shown, real-time sensor data is acquired from sensors, and timestamp records and device identifiers are fused to generate raw data records containing unique identifiers. If the sensor data lacks timestamp records or device identifiers, invalid data is discarded according to preset verification rules to obtain complete raw data records. A data fusion algorithm is used to process the raw data records, performing time-series alignment of the sensor data based on timestamp records to generate standardized data records. Standardized data records are associated with physical entities and bound to corresponding digital models through preset mapping rules to obtain data records bound to the digital models. If the data records bound to the digital models are inconsistent with the physical entity's state, anomaly detection is performed on the data records using a support vector machine algorithm to identify abnormal data records. Based on the anomaly detection results, the state parameters of the digital models are updated to generate traceable data records reflecting the process status. An index is built on the traceable data records using timestamp records and device identifiers to obtain a traceable dataset that supports fast querying.

[0032] Specifically, when extracting sensor data from the consistent process execution status, real-time data streams can be obtained from device sensors through an Industrial Internet of Things (IIoT) platform. For example, a temperature sensor collects data once per second, assuming a temperature value of 25.3°C is collected at a certain moment, and a pressure sensor collects 2.5 bar. This data is transmitted to a central server via the MQTT protocol. The server stores the data using a time-series database such as InfluxDB, automatically appending a timestamp (e.g., 2025-07-24T01:09:00.000Z) and a device identifier (e.g., DeviceID:A001). To ensure data integrity, the system verifies the collected data using the CRC32 algorithm to calculate the checksum. If the verification fails, a retransmission mechanism is triggered. When generating data records by combining timestamps and device identifiers, the system packages sensor data, timestamps, and device identifiers into a JSON format, such as {“device_id”:“A001”,“timestamp”:“2025-07-24T01:09:00.000Z”,“temperature”:25.3,“pressure”:2.5}, and sends it to the data processing module via the Kafka streaming platform. When binding physical entities to digital models, the system utilizes digital twin technology to map the real-time data of physical device A001 to its digital model. The JSON records are pushed to the digital twin platform using a REST API. The platform updates the model state using predefined mapping rules (such as temperature and the model's heat distribution function T(x,y,z)=25.3+0.1x). When generating traceable datasets, the system stores all records in a distributed database such as MongoDB and generates a hash value (such as SHA-256) for each record using blockchain technology (such as Hyperledger Fabric) and stores it on-chain to ensure data immutability. The analysis process includes anomaly detection, using the Z-score algorithm (formula: Z=(x-μ) / σ, where μ is the historical average of 24.8 and σ is 0.5) to determine whether a temperature of 25.3°C is abnormal (Z=1.0<2, no anomaly). If business correlation is required, the system can combine production plan data (such as order number ORD123) with sensor data to ensure that the correspondence between process execution and orders can be queried during traceability. This ultimately forms a dataset containing equipment, time, process parameters, and order information for subsequent quality analysis.

[0033] Step 5: For the traceable dataset, use blockchain technology to encrypt and store the mapping relationship between sensor data and process results, determine whether the data has been tampered with, and obtain the tamper-proof verification result.

[0034] like Figure 6As shown, data is collected by sensors, generating a mapping relationship between sensor data and process results. A hash algorithm is used to calculate the hash value of the mapping relationship, resulting in an initial hash record. This initial hash record is stored on the blockchain, and distributed ledger technology is used to generate an on-chain record, retrieving the stored hash value. The on-chain record is retrieved from the blockchain, and the stored hash value is extracted to obtain on-chain hash data. Data is collected again by sensors, generating a new mapping relationship. The same hash algorithm is used to calculate the hash value of the new mapping relationship, obtaining the current hash value. If the on-chain hash data matches the current hash value, the data is determined not to have been tampered with, and a data integrity verification result is obtained. If the on-chain hash data does not match the current hash value, a timestamp comparison is used to analyze the generation time of the on-chain record and the current data to determine the time point of tampering, obtaining a tampering verification result. Based on the tampering verification result, digital signature technology is used to encrypt and store the untampered data, generating a new on-chain record, and obtaining a tamper-proof storage result.

[0035] Specifically, to achieve traceable datasets and encrypt the mapping relationship between sensor data and process results using blockchain technology, and to determine whether the data has been tampered with, the following technical methods can be used. First, taking a temperature and humidity sensor as an example, assume that in an industrial scenario, the sensor collects temperature values ​​(e.g., 25.3°C) and humidity values ​​(e.g., 60.2%) every second, generating a data point with a timestamp of 2025-07-24 01:00:00. Combined with process results (e.g., product qualification rate 98.5%), a mapping relationship is formed: {"Time": "2025-07-24 01:00:00", "Temperature": 25.3, "Humidity": 60.2, "Quality Rate": 98.5}. This data is then hashed using the SHA-256 algorithm, for example, "8f4d9c2a...", serving as the digital fingerprint of the data. Next, this hash value, along with the data, is stored on the blockchain using Ethereum smart contracts. The smart contract defines a storage function `storeData(hash, data)` to write data and a hash to the blockchain, with a transaction ID such as "0x5b3e...". To ensure data encryption, the original data is encrypted using the AES-256 algorithm with a 32-byte random string as the key (e.g., "k9x...z2"). The encrypted ciphertext and hash are uploaded to the blockchain and stored in the IPFS distributed file system, obtaining a content identifier (e.g., "QmX...yZ"). To verify whether the data has been tampered with, the system periodically runs a verification program, reading the hash value and ciphertext from the blockchain, decrypting them, and recalculating the hash value. For example, reading data with a timestamp of 2025-07-24 01:00:00, decrypting it yields a temperature of 25.3°C, recalculating the hash value "8f4d9c2a...", and comparing it with the hash stored on the blockchain. If they match, the data has not been tampered with; if they do not match, such as tampering with it to 25.4°C causing the hash to become "7a2e...", then the data is determined to have been tampered with. The verification result is recorded as {"Time": "2025-07-24 01:00:00", "Status": "Untampered"}, and an event is triggered via smart contract to notify relevant business systems, such as the quality monitoring system, to automatically adjust process parameters (e.g., adjusting the temperature control threshold from 25.5°C to 25.2°C). Through the above process, data collection, storage, encryption, and verification form a closed loop, ensuring traceability and tamper-proof nature, with rigorous logic and full automation.

[0036] Step 6: Based on the anti-tampering verification results, extract historical process parameters and brake drum quality indicators from the traceable dataset, and use machine learning regression algorithm to analyze the correlation between parameters and quality to obtain an optimized parameter adjustment strategy.

[0037] like Figure 7As shown, historical process parameters and brake drum quality indicators are extracted from a traceable dataset to construct a structured dataset containing parameters and quality, resulting in an initial dataset. Data cleaning techniques are used to preprocess the initial dataset, removing missing and outlier values ​​to obtain a cleaned dataset. A linear regression algorithm is used to train the cleaned dataset, analyzing the correlation between process parameters and brake drum quality to obtain a regression model. If the prediction error of the regression model is below a preset threshold, process parameters that significantly affect quality are extracted from the regression model to obtain a set of key parameters; if the prediction error is above the preset threshold, a random forest algorithm is used to retrain the model, resulting in an updated regression model and set of key parameters. Based on the set of key parameters, the weighted influence of each parameter on brake drum quality is calculated, resulting in a parameter weight distribution. Using the parameter weight distribution, gradient descent is used to optimize the process parameter values, resulting in an optimized parameter adjustment strategy. Based on the optimized parameter adjustment strategy, a configuration file for process parameter adjustment is generated, outputting the adjusted parameter values.

[0038] Specifically, tamper-proof verification utilizes blockchain technology to check the integrity of the dataset, ensuring that the data has not been tampered with. Assuming the verification results show consistent data hash values, the traceable dataset is confirmed to be valid. The dataset contains historical production records of brake drums, extracting 1000 process parameter data points, including casting temperature (°C), cooling time (minutes), mold pressure (MPa), and quality indicators (such as hardness HB, wear resistance mm³, and weight kg). Example data: casting temperature 1500°C, cooling time 30 minutes, mold pressure 10MPa, corresponding to a hardness of 200HB, wear resistance 0.05mm³, and weight 15kg. A random forest regression algorithm is used to analyze the correlation between parameters and quality indicators. The scikit-learn library in Python is used, with n_estimators=100 and max_depth=10. The dataset is divided into an 80% training set and a 20% test set. After training the model, feature importance analysis shows that casting temperature has a 0.55% impact on hardness, cooling time has a 0.40% impact on wear resistance, and mold pressure has a 0.35% impact on weight. The model's mean squared error was 0.02, and the R² was 0.85, indicating a good model fit. Optimizing hyperparameters through grid search, adjusting the pouring temperature to 1520℃, cooling time to 28 minutes, and mold pressure to 12MPa, improved the predicted hardness to 205HB, reduced wear resistance to 0.045mm³, and stabilized the weight at 15.2kg. Based on the analysis, the optimization strategy is: prioritize adjusting the pouring temperature to increase hardness, shorten the cooling time to improve wear resistance, and moderately increase the mold pressure to control weight. This strategy is integrated into the production control system via automated scripts to adjust equipment parameters in real time, ensuring quality optimization.

[0039] Step 7: Based on the optimized parameter adjustment strategy, update the control parameters of the heat treatment furnace group in real time, distribute the strategy through the inter-equipment collaborative communication network, determine whether the adjusted parameters meet the production consistency requirements, and obtain the dynamically optimized process execution status.

[0040] like Figure 8 As shown, real-time process parameters are acquired from the heat treatment furnace group via a collaborative communication network and stored in a pre-established process database to obtain a real-time parameter set. Based on the real-time parameter set, a support vector machine algorithm is used to analyze parameter fluctuation characteristics and determine the changing trends of key process parameters. If the changing trend exceeds a preset threshold, a new parameter configuration is generated through a dynamic adjustment strategy, resulting in an optimized parameter set. The optimized parameter set is distributed to each device in the heat treatment furnace group through a parameter distribution mechanism, updating the device control parameters and obtaining the updated parameter execution status. The updated parameter execution status is then used, and a logistic regression algorithm is employed to determine whether the parameters meet production consistency requirements, yielding a consistency evaluation result. Based on the consistency evaluation result, if consistency is not met, the transmission priority of the collaborative communication network is adjusted, and the parameter distribution mechanism is optimized to obtain an improved distribution strategy. The optimized parameter set is redistributed using the improved distribution strategy to obtain a new process execution status and determine whether the dynamic optimization status is stable.

[0041] Specifically, real-time data from the heat treatment furnace group, such as furnace temperature (set to 1200°C, actual fluctuation range 1195-1205°C), atmosphere pressure (0.1MPa), and workpiece surface hardness (target value HRC50±2), is collected, and control parameters are optimized using a genetic algorithm. The algorithm initializes the population size to 100, iterates 50 times, has a crossover probability of 0.8, a mutation probability of 0.01, and the objective function is to minimize the weighted sum of temperature deviation and hardness deviation (weights of 0.6 and 0.4, respectively). The optimal parameter combination is calculated: heating power adjusted to 85kW, and holding time extended to 2.5 hours. The optimized parameters are distributed to each furnace via industrial Ethernet (transmission rate 100Mbps, latency less than 10ms), and the MQTT protocol is used to ensure parameter synchronization. After distribution, the system monitors the execution status of each furnace in real time, collects temperature and hardness data at 10-second intervals, and calculates consistency indicators: the standard deviation of temperature deviation must be less than 2°C, and the standard deviation of hardness deviation must be less than HRC1. Analysis showed that 90% of the furnace groups met the consistency requirements, while the remaining 10% had slightly higher deviations (temperature deviation 2.5°C) due to equipment aging. For the inconsistent equipment, the system automatically triggered secondary optimization, adjusting the heating power to 88kW and redistributing parameters. Dynamic monitoring showed that all furnace groups achieved consistency within 30 minutes, the process execution status stabilized, hardness deviation decreased to HRC0.8, and temperature deviation decreased to 1.8°C. The entire process was completed through edge computing nodes, requiring no manual intervention; the optimization algorithm and communication network worked together to ensure production consistency.

[0042] Step 8: Extract real-time sensor data from the dynamically optimized process execution status, update the parameter mapping of the digital model, determine whether the synchronization deviation between the model and the physical entity is within the preset range, and obtain the synchronization verification result.

[0043] like Figure 9 As shown, real-time sensor data is acquired from the process execution status. Temperature, pressure, and flow parameters are extracted at a preset frequency through the data acquisition module to obtain real-time sensor data. Based on the real-time sensor data, the digital model parameters are updated using a parameter mapping algorithm, and the model parameters are adjusted using a linear regression method to obtain the updated digital model parameters. The updated digital model parameters are compared with the physical entity state, and the synchronization deviation between the two is calculated using the Euclidean distance formula to obtain the synchronization deviation value. The Euclidean distance formula is: In the formula, Indicates the parameter values ​​of the digital model. Represents the state value of a physical entity.

[0044] If the synchronization deviation is less than a preset threshold, the synchronization is considered normal, and a synchronization verification result is obtained. If the synchronization deviation is greater than or equal to the preset threshold, a parameter adjustment mechanism is triggered, resulting in a synchronization anomaly signal. Based on the synchronization anomaly signal, the digital model parameters are optimized using a gradient descent algorithm. The synchronization deviation is reduced through iterative calculations, yielding optimized digital model parameters. Using the optimized digital model parameters, the synchronization deviation with the physical entity state is recalculated using the Euclidean distance formula to obtain a new synchronization deviation value. If the new synchronization deviation is less than the preset threshold, a synchronization verification result is output. If the new synchronization deviation is still greater than or equal to the preset threshold, deviation analysis data is recorded, resulting in the final verification result.

[0045] Specifically, when extracting real-time sensor data from the dynamically optimized process execution status, temperature sensor data can be collected at a frequency of 10 times per second through an industrial IoT platform. For example, the temperature value of a chemical reactor can be collected, assuming the current value is 85.3°C. The data is transmitted to a cloud database via the MQTT protocol and stored using the time-series database Influx DB to ensure efficient writing and querying of high-frequency data. Next, the parameter mapping of the digital model is updated. Assuming the digital model is a thermodynamic model of the reactor, it includes a parameter mapping table for temperature T, pressure P, and reaction rate k. Based on the collected 85.3°C, combined with the ideal gas law PV=nRT, the temperature parameter T in the model is updated, and the k value is updated through historical data regression analysis, assuming k=0.023mol / (L·s). Subsequently, the synchronization deviation between the model and the physical entity is determined, and a preset range of ±1°C temperature deviation is defined. The difference between the model's predicted temperature (calculated as 86.0°C based on the k value and the thermodynamic equation) and the actual 85.3°C is calculated, yielding a deviation of 0.7°C, which is less than 1°C and meets the synchronization requirement. Finally, the synchronization verification result is obtained. The system generates a boolean value of true by comparing the deviation, indicating that the synchronization is valid. The result is stored in a Redis cache and pushed to the monitoring system via a RESTful API, triggering subsequent process optimization adjustments, such as adjusting the heating power by 5kW to maintain temperature stability. The entire process is implemented through automated scripts, relying on Python's Paho-MQTT library for data acquisition, NumPy for parameter calculation, and the Flask framework for deviation analysis and result storage, ensuring rigorous logic and strong real-time performance.

[0046] Step 9: Based on the synchronization verification results, adjust the feedback control logic of the central data processing module, generate new process parameter synchronization instructions, and transmit them to the heat treatment furnace group to obtain the updated standardized parameter dataset.

[0047] like Figure 10As shown, deviation data is obtained from the synchronous verification results, and statistical analysis methods are used to determine the deviation range, providing a basis for parameter adjustment. Based on the deviation range, the feedback control logic of the central data processing module is adjusted, and the control parameters are updated using a PID algorithm to obtain optimized control logic. From the optimized control logic, process parameter synchronization instructions are generated, and the instruction content is formatted using a data encapsulation protocol to obtain a standardized instruction set. The standardized instruction set is transmitted to the heat treatment furnace group via a parameter synchronization transmission protocol. If data loss is detected during transmission, the instructions are resent to obtain a confirmed transmission status. Real-time feedback data is obtained from the heat treatment furnace group, and a data verification algorithm is used to determine the integrity of the feedback data, resulting in a verified feedback dataset. Based on the verified feedback dataset, the standardized parameter dataset is updated. If the parameter deviation exceeds a preset threshold, the control logic adjustment is retried to obtain the final parameter dataset. By comparing the final parameter dataset with the initial synchronous verification results, a consistency analysis method is used to determine the parameter synchronization effect, yielding the optimized results of the standardized process parameters.

[0048] Specifically, when adjusting the feedback control logic of the central data processing module based on the synchronous verification results, optimization can be achieved by analyzing the deviation between the real-time temperature data of the heat treatment furnace group and the target process parameters, and using a PID control algorithm. Assuming the synchronous verification results show a furnace group temperature deviation of ±5℃ and a target temperature of 800℃, the parameters of the PID algorithm are set to Kp=0.5, Ki=0.1, and Kd=0.05. By calculating the error e(t)=800-Tactual, the output control quantity u(t)=Kp·e(t)+Ki·∫e(t)dt+Kd·de(t) / dt is adjusted to generate a new feedback control logic. For example, when the actual temperature of a furnace is 795℃, the error e(t)=5, the cumulative error of the integral term is 20℃·s, and the rate of change of the derivative term is 0.2℃ / s, the calculated u(t)=0.5·5+0.1·20+0.05·0.2=4.51, and the output adjustment command increases the heating power by 4.51%. Subsequently, a new process parameter synchronization instruction is generated, including a temperature of 800℃, a holding time of 2 hours, and a heating rate of 5℃ / min. This instruction is encapsulated in JSON format as {"temp":800,"hold_time":7200,"rate":5} and transmitted to the edge computing nodes of the heat treatment furnace group via the MQTT protocol. Upon receiving the instruction, the furnace group updates its local control parameters and generates a standardized parameter dataset. This dataset includes fields such as furnace number, temperature, and timestamp, e.g., {"furnace_id":1,"temp":800,"timestamp":"2025-07-24T01:10:00Z"}. To ensure logical rigor, the system analyzes temperature stability by comparing the deviation between the updated dataset and historical data. If the standard deviation is less than 2℃, the parameters are considered valid; otherwise, a secondary adjustment is triggered. This process is automated using scripts, requiring no manual intervention, ensuring efficient and stable process parameter synchronization.

[0049] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0050] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A digital twin driven brake drum heat treatment process method, characterized by: Comprise: Step S1, obtain real-time process parameters from the heat treatment furnace group, collect and transmit to the central data processing module through the sensor network, and obtain a standardized parameter data set; Step S2, for the standardized parameter data set collected in step S1, a distributed consistency algorithm is used to coordinate the parameter synchronization of multiple heat treatment furnaces, and to determine whether the parameter deviation of each device exceeds the preset threshold, to obtain a synchronous adjustment instruction; Step S3, according to the synchronous adjustment instruction obtained in step S2, real-time update the temperature curve and atmosphere control parameters of each heat treatment furnace, realize collaborative adjustment through inter-device communication protocol, and obtain consistent process execution state; Step S4, further extract sensor data from the consistent process execution state, combine time stamp and device identifier, generate data record binding physical entity and digital model, and obtain traceable data set; Step S5, for the traceable data set, the mapping relationship between sensor data and process result is encrypted by using blockchain technology, and the anti-tamper verification result is obtained by judging whether the data is tampered with; Step S6, according to the anti-tamper verification result, extract historical process parameters and brake drum quality indicators from the traceable data set, analyze the correlation between parameters and quality by using machine learning regression algorithm, and obtain optimization parameter adjustment strategy; Step S7, according to the optimization parameter adjustment strategy, real-time update the control parameters of the heat treatment furnace group, distribute the strategy through the inter-device collaborative communication network, judge whether the adjusted parameters meet the production consistency requirements, and obtain the dynamically optimized process execution state; Step S8, extract real-time sensor data from the dynamically optimized process execution state, update the parameter mapping of the digital model, judge whether the synchronization deviation between the model and the physical entity is within the preset range, and obtain the synchronization verification result; Step S9, according to the synchronization verification result, adjust the feedback control logic of the central data processing module, generate new process parameter synchronization instruction, transmit to the heat treatment furnace group, and obtain updated standardized parameter data set.

2. The digital twin driven brake drum heat treatment process method according to claim 1, wherein: In step S1, the standardized parameter data set is obtained by collecting and transmitting to the central data processing module through the sensor network, comprising: Step S101, obtain real-time process parameters from the heat treatment furnace group through the sensor network, the process parameters include temperature data, atmosphere data and heating time, transmit to the central data processing module, and obtain the original parameter data set; Step S102, use data cleaning algorithm to preprocess the original parameter data set, remove abnormal values and missing values, and obtain the cleaned parameter data set; if the cleaned parameter data set meets the preset integrity threshold, normalize the temperature data, atmosphere data and heating time through standardization processing, and obtain the standardized parameter data set; Step S103, according to the standardized parameter data set, use K-means clustering algorithm to classify the furnace group state, and obtain the running mode set of process parameters; if a mode in the running mode set deviates from the preset process specification threshold, predict the parameter adjustment trend by using regression analysis algorithm, and obtain the adjustment parameter set; Step S104, according to the adjustment parameter set, generate process parameter optimization instructions for the furnace group, transmit to the heat treatment furnace group control system, get the optimized furnace group running state; Step S105, through real-time monitoring of the optimized furnace group running state, collect new process parameters, and cyclically update the standardized parameter data set.

3. The digital twin driven brake drum heat treatment process method according to claim 2, wherein: In step S2, for the standardized parameter data set, a distributed consistency algorithm is used to coordinate the parameter synchronization of multiple heat treatment furnaces, to determine whether the parameter deviation of each device exceeds the preset threshold, and to obtain a synchronization adjustment instruction, including: Step S201, obtain the running data of multiple heat treatment furnaces, and generate a standardized data set through standardization processing; Step S202, use a distributed consistency algorithm to extract each device parameter from the standardized data set and perform parameter synchronization operation; if any device parameter is detected to be inconsistent with the master node parameter during parameter synchronization, calculate the parameter deviation; and compare the parameter deviation with the preset threshold to determine whether the deviation exceeds the threshold range; Step S203, generate a synchronization instruction according to the deviation judgment result, the synchronization instruction contains specific adjustment instructions; Step S204, distribute the adjustment instruction to the corresponding heat treatment furnace through the distributed consistency algorithm, and perform parameter adjustment; Step S205, collect new parameters from the adjusted heat treatment furnace, verify the execution effect of the synchronization instruction, and generate a verification result.

4. The digital twin driven brake drum heat treatment process method of claim 3, wherein: In step S3, according to the synchronization adjustment instruction, real-time update the temperature curve and atmosphere control parameter of each heat treatment furnace, realize collaborative adjustment through inter-device communication protocol, obtain consistent process execution state, including: Step S301, obtain the real-time temperature curve and atmosphere control parameter of each heat treatment furnace through the inter-device communication protocol, and obtain a unified data set; If the temperature curve deviation in the data set exceeds the preset threshold, correct the heating power of each furnace through the synchronization adjustment algorithm to determine the consistent temperature curve; Step S302, according to the corrected temperature curve, adjust the oxygen content and pressure parameter of each furnace using the preset atmosphere control model to obtain optimized atmosphere control parameter; Step S303, transmit the optimized temperature curve and atmosphere control parameter to each heat treatment furnace through the inter-device communication protocol to obtain the collaborative execution control instruction; Step S304, if there is a deviation between the collaborative execution control instruction and the current furnace state, fine-tune the instruction through a feedback adjustment algorithm, and determine the consistent process execution state; Step S305, obtain the fine-tuned control instruction, distribute it to each heat treatment furnace through the inter-device communication protocol, and obtain the final process execution state.

5. The digital twin driven brake drum heat treatment process method according to claim 4, wherein: In step S4, extract sensor data from the consistent process execution state, combine time stamp and device identifier to generate data records binding physical entities and digital models, and obtain traceable data set, including: Step S401, obtain real-time collected sensor data from the sensor, fuse time stamp record and device identifier to generate original data record containing unique identifier; If the sensor data is missing time stamp record or device identifier, discard invalid data through the preset verification rule to obtain complete original data record; In step S402, the original data records are processed by using a data fusion algorithm, the sensor data is time-aligned based on the timestamp records, and standardized data records are generated; In step S403, the standardized data records are associated with the physical entity by using a preset mapping rule, are bound to the corresponding digital model, and data records bound to the digital model are obtained; If the data records bound to the digital model are inconsistent with the state of the physical entity, the data records are detected for abnormalities by using a support vector machine algorithm, and abnormal data records are determined; In step S404, according to the abnormal detection result, the state parameters of the digital model are updated, and traceable data records reflecting the process state are generated; In step S405, the traceable data records are indexed and constructed by using the timestamp records and the equipment identifier, and a traceable data set supporting fast query is obtained.

6. The digital twin driven brake drum heat treatment process method of claim 5, wherein: In step S5, for the traceable data set, the mapping relationship between the sensor data and the process result is encrypted by using a blockchain technology, it is judged whether the data is tampered with, and a tamper-proof verification result is obtained, including: In step S501, data is collected by using a sensor, a mapping relationship between the sensor data and the process result is generated, a hash value of the mapping relationship is calculated by using a hash algorithm, and an initial hash record is obtained; In step S502, the initial hash record is stored in a blockchain, a chain record is generated by using a distributed ledger technology, and a hash value stored in the chain is obtained; In step S503, the chain record is obtained from the blockchain, the stored hash value is extracted, and chain hash data is obtained; In step S504, data is collected again by using the sensor, a new mapping relationship is generated, a hash value of the new mapping relationship is calculated by using the same hash algorithm, and a current hash value is obtained; If the chain hash data is consistent with the current hash value, it is determined that the data has not been tampered with, and a data integrity verification result is obtained; If the chain hash data is inconsistent with the current hash value, the generation time of the chain record and the current data is compared and analyzed by using the timestamp, the time point of tampering is judged, and a tampering verification result is obtained; In step S505, according to the tampering verification result, the un-tampered data is encrypted and stored by using a digital signature technology, a new chain record is generated, and a tamper-proof storage result is obtained.

7. The digital twin driven brake drum heat treatment process method of claim 6, wherein: In step S6, according to the tamper-proof verification result, historical process parameters and brake drum quality indicators are extracted from the traceable data set, the relevance of the parameters and the quality is analyzed by using a machine learning regression algorithm, and an optimization parameter adjustment strategy is obtained, including: In step S601, historical process parameters and brake drum quality indicators are extracted from the traceable data set, a structured data set containing the parameters and the quality is constructed, and an initial data set is obtained; In step S602, the initial data set is preprocessed by using a data cleaning technology, missing values and abnormal values are removed, and a cleaned data set is obtained; In step S603, the cleaned data set is trained by using a linear regression algorithm, the relevance of the process parameters and the brake drum quality is analyzed, and a regression model is obtained; If the prediction error of the regression model is lower than a preset threshold, the process parameters significantly affecting the quality are extracted from the regression model, and a key parameter set is obtained; If the prediction error is higher than the preset threshold, the random forest algorithm is retrained to obtain an updated regression model and a set of key parameters; Step S604, according to the set of key parameters, the weight influence of each parameter on the brake drum mass is calculated to obtain the parameter weight distribution; Step S605, through the parameter weight distribution, the gradient descent method is used to optimize the process parameter value to obtain the optimization parameter adjustment strategy; Step S606, according to the optimization parameter adjustment strategy, the configuration file of process parameter adjustment is generated, and the adjusted parameter value is output.

8. The digital twin driven brake drum heat treatment process method of claim 7, wherein: In step S7, the control parameters of the heat treatment furnace group are updated in real time according to the optimization parameter adjustment strategy, the strategy is distributed through the inter-device collaborative communication network, it is judged whether the adjusted parameters meet the production consistency requirements, and the dynamic optimization process execution state is obtained, including: Step S701, real-time process parameters are obtained from the heat treatment furnace group through the collaborative communication network and stored in the pre-established process database to obtain a set of real-time parameters; Step S702, according to the set of real-time parameters, the support vector machine algorithm is used to analyze the parameter fluctuation characteristics, and the change trend of the key process parameters is determined; If the change trend exceeds the preset threshold, a new parameter configuration is generated through the dynamic adjustment strategy to obtain an optimized parameter set; Step S703, the optimized parameter set is distributed to each device of the heat treatment furnace group through the parameter distribution mechanism, the device control parameters are updated, and the updated parameter execution state is obtained; Step S704, the updated parameter execution state is obtained, the parameter is judged whether it meets the production consistency requirements by using the logistic regression algorithm, and the consistency evaluation result is obtained; Step S705, according to the consistency evaluation result, if the consistency is not up to standard, the transmission priority of the collaborative communication network is adjusted, the parameter distribution mechanism is optimized, and the improved distribution strategy is obtained; Step S706, the optimized parameter set is redistributed through the improved distribution strategy, the new process execution state is obtained, and it is judged whether the dynamic optimization state is stable.

9. The digital twin driven brake drum heat treatment process method of claim 8, wherein: In step S8, real-time sensor data is extracted from the dynamic optimization process execution state, the parameter mapping of the digital model is updated, it is judged whether the synchronization deviation between the model and the physical entity is within the preset range, the synchronization verification result is obtained, including: Step S801, real-time sensor data is obtained from the process execution state, temperature, pressure and flow parameters are extracted at a preset frequency through a data acquisition module to obtain real-time sensor data; Step S802, according to the real-time sensor data, the parameter mapping algorithm is used to update the digital model parameters, the model parameters are adjusted through the linear regression method, and the updated digital model parameters are obtained; Step S803, the updated digital model parameters are compared with the physical entity state, the Euclidean distance formula is used to calculate the synchronization deviation between them, and the synchronization deviation value is obtained, wherein the Euclidean distance formula is: wherein denotes a digital model parameter value, denotes a physical entity state value; If the synchronization deviation value is less than the preset range threshold, it is determined that the synchronization state is normal, and the synchronization verification result is obtained; If the synchronization deviation value is greater than or equal to the preset range threshold, the parameter adjustment mechanism is triggered, and the synchronization exception signal is obtained; Step S804, according to the synchronization exception signal, the gradient descent algorithm is used to optimize the digital model parameters, the synchronization deviation is reduced through iterative calculation, and the optimized digital model parameters are obtained; Step S805, the synchronization deviation with the physical entity state is recalculated through the optimized digital model parameters, and a new synchronization deviation value is obtained by using the Euclidean distance formula; If the new synchronization deviation value is less than the preset range threshold, the synchronization verification result is output; If the new synchronization deviation value is still greater than or equal to the preset range threshold, the deviation analysis data is recorded, and the final verification result is obtained.

10. The digital twin driven brake drum heat treatment process method of claim 9, wherein: In step S9, according to the synchronization verification result, the feedback control logic of the central data processing module is adjusted, a new process parameter synchronization instruction is generated, and is transmitted to the heat treatment furnace group, so that an updated standardized parameter data set is obtained, including: Step S901, deviation data is obtained from the synchronization verification result, a statistical analysis method is used to determine the deviation value range, and parameter adjustment basis is obtained; Step S902, according to the deviation value range, the feedback control logic of the central data processing module is adjusted, the control parameters are updated by using the PID algorithm, and the optimized control logic is obtained; Step S903, process parameter synchronization instructions are generated from the optimized control logic, the instruction content is formatted by using a data encapsulation protocol, and a standardized instruction set is obtained; Step S904, the standardized instruction set is transmitted to the heat treatment furnace group by using a parameter synchronization transmission protocol, if data loss is detected during transmission, the instruction is resent, and a confirmed transmission state is obtained; Step S905, real-time feedback data is obtained from the heat treatment furnace group, a data verification algorithm is used to judge the integrity of the feedback data, and a verified feedback data set is obtained; Step S906, according to the verified feedback data set, the standardized parameter data set is updated, if the parameter deviation exceeds the preset threshold, the control logic adjustment is retriggered, and the final parameter data set is obtained; Step S907, by comparing the final parameter data set with the initial synchronization verification result, a consistency analysis method is used to judge the parameter synchronization effect, and the optimization result of the standardized process parameters is obtained.

Images