Multi-unit thermodynamic system real-time data interface scheduling method and system

Abstract

The invention discloses a real-time data interface scheduling method and system for a multi-unit thermodynamic system, and belongs to the technical field of industrial internet scheduling, and the scheduling method comprises the following specific steps: I, collecting and processing the operation data of each unit in real time, and analyzing the local entropy change rate and change trend of each unit according to the processed operation data; according to the method, the forebodes of system operation disorder or control failure can be accurately captured, so that the system scheduling has good adaptivity, the data and system state depth correlation modeling capability is improved, the multi-element dynamic adjustment of the data flow priority is realized, the resource utilization efficiency is improved, and the data transmission efficiency and the aggregation compression capability are improved; according to the method, super-parallel and high-speed synchronous transmission can be realized, the optimal energy consumption constraint capability of path scheduling can be kept, the anti-interference capability of the system is improved, defense intelligence is enhanced, and the availability of a control instruction channel is effectively guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial internet scheduling, in particular to a multi-unit thermal system real-time data interface scheduling method and system. BACKGROUND

[0002] With the development of energy equipment intelligentization and multi-unit collaborative control, thermal systems are evolving towards high complexity, high concurrency and high reliability. Under this background, the exchange of data between different units is becoming more and more intensive, and various real-time state data, control instructions, early warning information, etc. need to be accurately synchronized through the interface layer at sub-second level. However, due to the limitation of traditional static scheduling mechanism, the interface scheduling efficiency and robustness cannot adapt to the data load fluctuation caused by the dynamic change of unit thermal cycle, which easily causes data congestion, link congestion and delayed transmission of key control information, affecting the stable operation of the system. In addition, the current interface layer processing mechanism for sudden failures, malicious attacks and cross-unit data synchronization is still relatively primitive, lacking deep perception and response ability to the disorder trend of system state and abnormal risk evolution process.

[0003] The existing multi-unit thermal system real-time data interface scheduling method and system cannot accurately capture the signs of system operation confusion or control failure, reduce the adaptability of system scheduling, and the data and system state deep association modeling capability is poor, which cannot realize the multi-element dynamic adjustment of data flow priority, reduces the resource utilization efficiency; in addition, the existing multi-unit thermal system real-time data interface scheduling method and system cannot realize super-parallel, high-speed synchronous transmission, cannot guarantee that the path scheduling maintains the optimal energy consumption constraint ability, reduces the system anti-interference ability, and cannot guarantee the availability of control instruction channel; therefore, we propose a multi-unit thermal system real-time data interface scheduling method and system. SUMMARY

[0004] The purpose of the present application is to solve the defects in the prior art and provide a multi-unit thermal system real-time data interface scheduling method and system.

[0005] In order to achieve the above purpose, the technical scheme adopted by the present application is as follows: 1. A multi-unit thermal system real-time data interface scheduling method, the scheduling method comprising the following steps: Ⅰ. Real-time acquisition and processing of operation data of each unit, analysis of local entropy change rate and trend of each unit according to the processed operation data of each unit; Ⅱ. Dynamically adjusting the scheduling priority of different data streams according to the local entropy change rate and trend of each unit; Ⅲ. Constructing a cross-unit virtual data channel according to the data stream scheduling priority, real-time analyzing the interface failure probability, and performing virtual isolation test; IV. According to the test results, the bandwidth allocation of each interface is dynamically adjusted, and the data stream instructions are synchronously transmitted among multiple units, while the scheduling path conflicts are resolved in real time; V. The total entropy of the multi-unit thermal system is detected in real time, and the data path structure between the interfaces is dynamically reconstructed, while the abnormal data stream is identified in real time during the data transmission process; VI. After the scheduling action is executed, the performance indicators in the current scheduling process are collected, and the next round of scheduling strategy is optimized.

[0006] As a further scheme of the present application, the specific steps of the real-time collection and processing of the operation data of each unit in step I are as follows: S1.1: The multi-unit thermal system collects the operation data of each unit in real time through various sensors, and eliminates the high-frequency noise in each type of operation data by using the moving average method, and then selects the median of each type of smoothed operation data as the dynamic reference value; S1.2: According to the dynamic reference value, the deviation rate of each type of operation data relative to the median is calculated, and then according to the historical data fluctuation characteristics, the dynamic deviation threshold of each data transmission channel is dynamically set by using the sliding standard deviation, if the deviation rate is higher than the current dynamic deviation threshold, the corresponding operation data is marked as an abnormal value, otherwise, it is marked as a normal value; S1.3: The number and position of missing values and abnormal values in each type of operation data are counted, and the missing values and abnormal values are corrected by using the neighborhood mean interpolation method, and the corrected operation data of each type is standardized to unify the dimensions of each type of operation data, and according to the preset rule, the unit format of each type of operation data of each unit is unified, and the Min-Max normalization processing is used to unify each type of operation data to the interval [0, 1].

[0007] As a further scheme of the present application, the specific steps of analyzing the local entropy rate and change trend of each unit in step I are as follows: S2.1: Extract the variable time series of each unit in the current thermal cycle from the processed operation data, and calculate the local entropy rate of the multi-unit thermal system in different time periods of the current thermal cycle based on the extracted variable time series of each unit; S2.2: Smooth filtering is performed on the calculated local entropy rate of each group to eliminate the interference of instantaneous disturbance on the corresponding local entropy rate, and then based on the smoothed local entropy rate, a state entropy value model of the multi-unit thermal system is constructed, and the global state entropy value of the multi-unit thermal system is obtained in real time through the state entropy value model; S2.3: Real-time monitoring of the entropy value of the multi-unit thermal system at different times, and calculation of the change in the entropy value of the multi-unit thermal system at the current time relative to the previous time, so as to obtain the rate of change of the entropy value trend of the multi-unit thermal system. At the same time, based on historical experience and simulation results, the critical entropy value is set. S2.4: If the current entropy trend change rate is higher than the critical entropy value, it indicates that the current multi-unit thermal system is in a state of operational chaos and has a risk of control failure. Based on the local entropy change rate and global state entropy value of the current multi-unit thermal system, the corresponding scheduling feedback factor of each unit in the current multi-unit thermal system is calculated to obtain the current impact of each unit on the entropy increase of the multi-unit thermal system.

[0008] As a further aspect of the present invention, the specific calculation formula for the local entropy change rate in S2.1 is as follows: In the formula, Representing the The unit at time The rate of change of local entropy; Representing the The unit at time The heat output; Representing the The unit at time Heat input; Representing the The unit at time The effective thermal temperature; The specific calculation formula for the smoothing filter described in S2.2 is as follows: In the formula, Representing the The unit at time The smoothed rate of local entropy change; Represents the smoothing coefficient; Representing the The unit at time The smoothed rate of local entropy change; The specific form of the state entropy model described in S2.2 is as follows: In the formula, Represents a multi-unit thermal system at any time The global state entropy value; Represents the total number of generating units participating in the scheduling; Representing the Weighting factors for individual generating units; The specific formula for calculating the rate of change of entropy value trend mentioned in S2.3 is as follows: In the formula, represents the change of the global state entropy value of the multi-unit thermal system at the time relative to the previous time; represents the global state entropy value of the multi-unit thermal system at the time The specific calculation formula of the scheduling feedback factor S2.4 is as follows: In the formula, represents the scheduling feedback factor of the i th unit to the system entropy change; represents a small constant, which is used to avoid division by zero.

[0009] As a further scheme of the present application, the specific steps of dynamically adjusting the scheduling priority of different data streams in step II are as follows: S3.1: According to the normalized running data of each type, the data stream set of each unit and its sub-module is established, and each data stream set is labeled according to the preset unit number. Then, according to the data stream set corresponding to each unit and the current smoothed local entropy change rate, an "entropy value-data stream" correlation matrix is constructed to obtain the entropy correlation weight between each unit and each data stream set in real time; S3.2: The entropy correlation weights are aggregated, the correlation strength of each data stream set and the current multi-unit thermal system entropy is obtained, and the correlation strength of each data stream set obtained is normalized. Then, the current entropy value trend change rate is monitored in real time; S3.3: If the current entropy value trend change rate is lower than the critical entropy value, the current scheduling priority factor of each data stream is calculated based on the normalized correlation strength of each data stream set and the basic service weight of the corresponding data stream. Then, the scheduling priority factors are arranged in descending order to form a scheduling queue. After the arrangement is completed, the data streams are taken out from the scheduling queue in turn to perform bandwidth allocation and transmission scheduling; S3.4: If the current entropy value trend change rate is higher than the critical entropy value, the current scheduling priority factor of each data stream is calculated based on the normalized correlation strength of each data stream set and the scheduling feedback factor. Then, the scheduling priority factors are arranged in descending order to form a scheduling queue. After the arrangement is completed, the data streams are taken out from the scheduling queue in turn to perform bandwidth allocation and transmission scheduling; S3.5: Based on the preset update time period, the latest entropy change rate is extracted regularly, and it is checked whether there are new or invalid data streams. According to the latest entropy change rate and the checking result, the scheduling priority factor of each data stream is dynamically updated.

[0010] ​​As a further scheme of the present application, the specific steps of constructing the cross-machine group virtual data channel according to the scheduling priority of each data stream in step III are as follows: S4.1: According to the current data stream scheduling priority arrangement result, the data stream set is extracted in turn from high to low, and the corresponding source machine group and target machine group information is recorded, and a data path mapping table is established; S4.2: Collect the sampling time stamp of each data stream and the position information of the corresponding source machine group in the logical topology of the current multi-machine group heat system, and add the collected sampling time stamp and source machine group position information to the corresponding data stream set; S4.3: Calculate the space-time similarity between each data stream set respectively, if the space-time similarity is higher than the preset similarity threshold, then merge the corresponding data stream set to construct a superframe data packet, and then the multi-machine group heat system registers a virtual channel for each superframe data packet in the interface layer; S4.4: After registration, according to the shortest delay strategy, find the physical path that meets the transmission constraint in the interface mapping matrix, and based on the search result, bind the virtual channel and the physical path obtained by searching, allocate transmission time slots for each virtual channel and start channel transmission; S4.5: Based on the preset channel update period, periodically update the space-time similarity between each data stream set and the logical bandwidth demand of each virtual channel, if the space-time correlation degree decreases or the bandwidth demand is higher than the preset threshold, then automatically cancel the mapping relationship when the old virtual channel data is completed transmission, release the physical resources, and dynamically construct a new virtual channel.

[0011] As a further scheme of the present application, it comprises a collection and processing module, an entropy change analysis module, a state evaluation module, a dynamic priority module, a channel construction module, a prediction simulation module, a conflict resolution module, a safety control module and a feedback control module; The collection and processing module is used for high-frequency collection and processing of real-time operation data from the data interface of each heat engine; The entropy change analysis module calculates the local entropy change rate of each machine group according to the processed operation data; The state evaluation module is used for integrating the local entropy change rate of each machine group, and evaluating the current multi-machine group heat system operation state in real time according to the integration result; The dynamic priority module is used for generating the scheduling priority factor of each data stream according to the real-time current multi-machine group heat system entropy weight distribution, and dynamically adjusting the scheduling order of each data stream; The channel construction module is used for constructing a reconfigurable virtual transmission channel in the interface layer, and sequentially performing data transmission according to the scheduling order of each data stream; The prediction simulation module monitors the interface layer signal in real time during data transmission, predicts each communication anomaly or fault, and performs sandbox isolation test on the potential fault interface; The conflict resolution module dynamically eliminates multi-path data conflicts and optimizes the scheduling routing structure according to the detection results of each virtual channel; The security control module detects abnormal data transmitted by each virtual channel in real time, and traps and isolates each abnormal data identified; The feedback control module is used for regularly analyzing system operation data, entropy state trend and channel load, and optimizing the next round of scheduling strategy based on the analysis results.

[0012] As a further scheme of the present application, the specific steps of the conflict resolution module dynamically eliminating multi-path data conflicts and optimizing the scheduling routing structure are as follows: S5.1: Each superframe data packet transmitted in real time in each virtual channel is converted by digital-quantum state encoding, the content of each superframe data packet is encoded into a corresponding quantum bit sequence, and then a microwave resonant cavity is used to establish a multi-node quantum entanglement channel between each virtual channel in the interface layer; S5.2: For all virtual channels between each unit, a path Hamiltonian model is constructed respectively, and then each path Hamiltonian model is converted into a corresponding QUBO model, and all virtual channels are initialized as uniformly distributed quantum superposition states; S5.3: Set the total evolution time, divide it into multiple equally spaced time steps, and use the Schrödinger equation to update the time evolution operator at each time step in real time to obtain the quantum state change rule of each unit at each time step, and then update the quantum state of each unit step by step using the time evolution operator; S5.4: During the evolution of each time step, the tunneling strength is updated in real time based on the tunneling strength control mechanism, and the total Hamiltonian of the corresponding time step is adjusted according to the corresponding point tunneling strength, and the evolution is stopped when the total evolution time is reached; S5.5: Measure the current quantum state, obtain the probability amplitude of each virtual channel combination, and take the virtual channel combination with the highest probability amplitude as the optimal solution of the virtual channel combination between the corresponding units, and then map the virtual channel combination to the virtual channel routing table and configure the bandwidth resources according to the node relationship.

[0013] As a further scheme of the present application, the specific calculation formula of the digital-quantum state encoding conversion in S5.1 is as follows: In the formula, represents the th quantum bit after mapping; ; represents the content of the th superframe data packet in the original data stream; and respectively represent complex probability amplitudes, and respectively represent standard basis states in quantum computation, representing logical 0 and 1; S5.2 The specific form of the quantum superposition state is as follows: In the formula, represents an initial quantum superposition state; represents the total number of optional virtual channels; represents each virtual channel combination; represents the quantum basis state corresponding to each virtual channel combination.

[0014] As a further scheme of the present application, the specific steps of the security control module for trapping and isolating each abnormal data identified are as follows: S6.1: Real-time monitoring of transmission data of each interface, and extracting flow rate, connection frequency and source IP distribution of each traffic feature from the monitored transmission data, while calculating the flow entropy degree of the current network connection state according to the extracted each traffic feature, if the flow entropy degree drop value of adjacent time is more than the preset abnormal threshold, it is judged as the DDoS behavior of botnet; S6.2: Based on historical protection experience, manual setting and expert knowledge, set multiple groups of abnormal behavior characteristics to establish corresponding antigen characteristic fingerprints, and then based on each antigen characteristic fingerprint, initialize multiple groups of candidate antibodies to form an immune population, and then calculate the recognition matching degree between each candidate antibody in the immune population and the antigen characteristic fingerprint; S6.3: Select each candidate antibody with a recognition matching degree higher than a preset threshold, and repeatedly perform cloning, mutation and screening operations on each candidate antibody selected, until the recognition matching degree change value of each candidate antibody in the immune population converges to within a preset termination threshold, and based on the final immune population, a corresponding response strategy model is established; S6.4: The response strategy model analyzes whether each real-time transmission data is an attack behavior in real time, if it is judged as an attack behavior, the virtual bandwidth bait is activated, and the attackable target is simulated through the bandwidth bait to attract malicious data flow, while the occupied bandwidth used for attack in the real interface is released, when the response strategy model detects a new type of attack or antibody recognition fails, the corresponding transmission data is automatically added to the library as a new antigen sample, and the immune population of the response strategy model is updated.

[0015] Compared with the prior art, the present application has the following advantages: 1、The present application collects and processes the operation data of each unit, extracts the time series of each variable of the thermal cycle, calculates and smooths the local entropy rate of change, constructs a global state entropy model, then monitors the entropy value trend in real time based on the global state entropy model, judges whether the system has operation risk according to the entropy trend and the critical threshold, calculates the scheduling feedback factor combining the local entropy and the global entropy, constructs the "entropy value-data flow" correlation matrix, evaluates the influence of each data flow on the system entropy increase, at the same time, dynamically calculates the scheduling priority combining the business weight or scheduling feedback factor of the data flow, constructs the scheduling queue, executes bandwidth allocation and scheduling transmission, at the same time, calculates the space-time similarity based on the data flow sampling time and the source unit position information, automatically aggregates into superframe data packets and registers the virtual channel, completes the physical path mapping and time slot allocation through the interface layer, realizes efficient compression transmission of data, can accurately capture the premonition of system operation confusion or control failure, makes the system scheduling have good self-adaptability, improves the data and system state deep correlation modeling capability, realizes the multi-element dynamic adjustment of data flow priority, improves the resource utilization efficiency, improves the data transmission efficiency and aggregation compression capability.

[0016] 2、The present application encodes the superframe data transmitted in the virtual channel into a sequence of qubits, and constructs a multi-node quantum entanglement channel in the interface layer, constructs a path Hamiltonian model for each unit virtual channel and converts it into a QUBO model, uses quantum annealing process to solve the optimal routing scheme, at the same time, dynamically updates the quantum state through the time evolution operator and the tunneling strength, to obtain the optimal channel combination and configure the bandwidth, and synchronously monitor the interface traffic, extract the features and calculate the traffic entropy, identify the abnormal downward trend to detect DDoS attack, use immune algorithm to construct antigen-antibody recognition mechanism, get high matching immune strategy through candidate antibody population evolution, start virtual decoy transfer flow for attack behavior, release real link resources, when new attack is encountered, mark it as a new antigen sample, automatically update the immune library and response strategy, can realize super-parallel, high-speed synchronous transmission, can make the path scheduling maintain the optimal energy consumption constraint ability, improve the system anti-interference ability, at the same time, strengthen the defense intelligence, effectively guarantee the availability of control instruction channel. BRIEF DESCRIPTION OF DRAWINGS

[0017] The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application, and do not limit the application.

[0018] Figure 1 A flow chart of a multi-unit thermal system real-time data interface scheduling method according to the present application is shown in the figure. Figure 2 A system block diagram of a multi-unit thermal system real-time data interface scheduling system according to the present application is shown in the figure. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application.

[0020] Embodiment 1 With reference to Figure 1 The embodiment discloses a real-time data interface scheduling method of a multi-unit thermal system, and the specific steps are as follows: Real-time collection and processing of operation data of each unit, analysis of local entropy change rate and change trend of each unit according to the processed operation data.

[0021] Specifically, the multi-unit thermal system classifies and collects the operation data of each unit in real time through various sensors, and eliminates high-frequency noise in each type of operation data by using a moving average method. Then, the median of each type of smoothed operation data is selected as a dynamic reference value. According to each dynamic reference value, the deviation rate of each type of operation data relative to the median is calculated. Then, according to the historical data fluctuation characteristics, the dynamic deviation threshold of each data transmission channel is dynamically set by using a sliding standard deviation. If the deviation rate is higher than the current dynamic deviation threshold, the corresponding operation data is marked as an abnormal value, otherwise, it is marked as a normal value. The number and position of missing values and abnormal values in each type of operation data are counted, and the missing values and abnormal values are corrected by using a neighborhood mean interpolation method. The corrected operation data of each type is standardized to unify the dimensions of each type of operation data. At the same time, according to a preset rule, the units of each type of operation data of each unit are unified, and the Min-Max normalization processing is used to unify each type of operation data to the interval [0, 1].

[0022] Specifically, the time series of each variable for each unit in the current thermodynamic cycle is extracted from the processed operating data. Based on the extracted time series of each unit, the local entropy change rate of the multi-unit thermodynamic system in different time periods of the current thermodynamic cycle is calculated. The calculated local entropy change rates are smoothed and filtered to eliminate the interference of instantaneous disturbances on the corresponding local entropy change rates. Then, based on the smoothed local entropy change rates, a state entropy model of the multi-unit thermodynamic system is constructed. The global state entropy value of the multi-unit thermodynamic system is obtained in real time through the state entropy model, and the multi-unit thermodynamic system is monitored in real time at different times. The entropy value is calculated, and the change in the entropy value of the multi-unit thermal system at the current moment relative to the previous moment is also calculated to obtain the entropy trend rate of change of the multi-unit thermal system. At the same time, based on historical experience and operation simulation results, a critical entropy value is set. If the current entropy trend rate of change is higher than the critical entropy value, it indicates that the current multi-unit thermal system is in a state of operational chaos and has a risk of control failure. Based on the local entropy change rate and the global state entropy value of the current multi-unit thermal system, the scheduling feedback factor corresponding to each unit in the current multi-unit thermal system is calculated to obtain the current impact of each unit on the entropy increase of the multi-unit thermal system.

[0023] In addition, it should be further explained that the specific operating data of each unit includes: thermal parameters, structural status parameters, interface communication parameters, and energy consumption indicators; The time series of variables in the thermodynamic cycle of a multi-unit thermal system specifically include: heat input, heat output, working fluid temperature, pressure, and flow rate; The specific calculation formula for the moving average method is as follows: In the formula, Representing the Each data channel at time Smoothed running data; Represents the width of the sliding window; Represents the current point in time; Representing the At the [time]th moment The raw operating data of each data channel; The specific formula for calculating the deviation rate is as follows: In the formula, Representing the Each data channel at time The relative deviation rate of the operating data; Representing the Each data channel at time The sliding median of the running data; It represents a very small positive number and is used to prevent the denominator from being zero; The specific calculation formula of the dynamic deviation threshold is as follows: In the formula, represents the sliding standard deviation of the first data channel with time as the center; represents the dynamic deviation threshold of the first data channel; represents the smoothed running data of the first data channel at time ; represents a coefficient for controlling sensitivity, and the value range is [2, 4]; The specific calculation formula of the neighborhood mean interpolation method is as follows: In the formula, represents the running data of the first data channel at time after correction; represents a set of effective sampling time points adjacent to time ; represents the number of the set of effective sampling time points; represents the smoothed running data in the neighborhood range ; The specific calculation formula of the local entropy change rate is as follows: In the formula, represents the local entropy change rate of the first unit at time ; represents the heat output of the first unit at time ; represents the heat input of the first unit at time ; represents the effective thermodynamic temperature of the first unit at time ; The specific calculation formula of the smoothing filter is as follows: In the formula, represents the smoothed local entropy change rate of the first unit at time ; represents a smoothing coefficient; represents the first unit at time Smoothed local entropy rate of change; The specific form of the state entropy value model is as follows: In the formula, represents the global state entropy value of the multi-unit thermal system at time ; represents the total number of units participating in scheduling; represents the weight factor of the th unit; The specific calculation formula of the entropy value trend rate of change is as follows: In the formula, represents the change amount of the global state entropy value of the multi-unit thermal system at time relative to the previous time; represents the global state entropy value of the multi-unit thermal system at time ; The specific calculation formula of the scheduling feedback factor is as follows: In the formula, represents the scheduling feedback factor of the th unit to the system entropy change; represents a small constant, used to avoid division by zero.

[0024] According to the local entropy rate of change and the change trend of each unit, the scheduling priority of different data streams is dynamically adjusted.

[0025] Specifically, according to the normalized operation data of each type, the data flow set of each unit and its sub-module is established, and each data flow set is labeled according to the preset unit number. Then, according to the data flow set corresponding to each unit and the current smoothed local entropy change rate, an "entropy-data flow" correlation matrix is constructed to obtain the entropy correlation weight between each unit and each data flow set in real time. The entropy correlation weights are aggregated to obtain the correlation strength of each data flow set and the current multi-unit thermal system entropy. The correlation strength of each data flow set is normalized, and the current entropy value trend change rate is monitored in real time. If the current entropy value trend change rate is lower than the critical entropy value, the current scheduling priority factor of each data flow is calculated based on the normalized correlation strength of each data flow set and the basic business weight of the corresponding data flow. Then, the scheduling priority factor is arranged in descending order to form a scheduling queue. After the arrangement is completed, the data flow is taken out from the scheduling queue in turn to perform bandwidth allocation and transmission scheduling. If the current entropy value trend change rate is higher than the critical entropy value, the current scheduling priority factor of each data flow is calculated based on the normalized correlation strength of each data flow set and the scheduling feedback factor. Then, the scheduling priority factor is arranged in descending order to form a scheduling queue. After the arrangement is completed, the data flow is taken out from the scheduling queue in turn to perform bandwidth allocation and transmission scheduling. Based on the preset update time period, the latest entropy change rate is extracted regularly, and it is checked whether there is a new or invalid data flow. According to the latest entropy change rate and the checking result, the scheduling priority factor of each data flow is dynamically updated.

[0026] According to the scheduling priority of each data flow, a corresponding cross-unit virtual data channel is constructed, and the interface failure probability is analyzed in real time, and virtual isolation testing is performed.

[0027] Specifically, according to the current data flow scheduling priority arrangement result, the data flow set is extracted from high to low in turn, and the corresponding source unit and target unit information is recorded, and a data path mapping table is established, the sampling time stamp of each data flow is collected, and the position information of the corresponding source unit in the logical topology of the current multi-unit heat system is collected, and the collected sampling time stamp and source unit position information are added to the corresponding data flow set, the space-time similarity between each data flow set is calculated respectively, if the space-time similarity is higher than the preset similarity threshold, the corresponding data flow set is merged to construct a superframe data packet, then the multi-unit heat system registers a virtual channel for each superframe data packet in the interface layer, after registration is completed, according to the shortest delay strategy, the physical path meeting the transmission constraint is searched in the interface mapping matrix, and based on the search result, the virtual channel and the searched physical path are bound, the transmission time slot is allocated for each virtual channel and the channel transmission is started, based on the preset channel update period, the space-time similarity between each data flow set and the logical bandwidth demand of each virtual channel are updated regularly, if the space-time correlation degree decreases or the bandwidth demand is higher than the preset threshold, the mapping relationship is automatically deregistered after the old virtual channel data is completed transmission, the physical resources are released, and a new virtual channel is dynamically constructed.

[0028] According to the test results, the bandwidth allocation of each interface is dynamically adjusted, and the data flow instructions are synchronously transmitted between multiple units, while the scheduling path conflicts are resolved in real time.

[0029] The total entropy of the multi-unit heat system is detected in real time, and the data path structure between the interfaces is dynamically reconstructed, and in the data transmission process, the abnormal data flow is identified in real time.

[0030] After the scheduling action is completed, the performance indicators in the current scheduling process are collected, and the next round of scheduling strategy is optimized.

[0031] Embodiment 2 Referring to Figure 2 The embodiment discloses a multi-unit heat system real-time data interface scheduling system, which comprises an acquisition and processing module, an entropy change analysis module, a state evaluation module, a dynamic priority module, a channel construction module, a prediction simulation module, a conflict resolution module, a safety control module and a feedback control module.

[0032] The acquisition and processing module is used for high-frequency acquisition and processing of real-time operation data from the data interfaces of each heat unit; the entropy change analysis module calculates the local entropy change rate of each unit according to the processed operation data.

[0033] The state evaluation module is used for integrating the local entropy change rates of each unit and evaluating the current multi-unit heat system operation state in real time according to the integration result; the dynamic priority module is used for generating a scheduling priority factor of each data flow according to the real-time current multi-unit heat system entropy weight distribution, and dynamically adjusting the scheduling order of each data flow.

[0034] The channel construction module is used to construct reconfigurable virtual transmission channels at the interface layer, and sequentially perform data transmission according to the data flow scheduling sequence; the prediction simulation module monitors the interface layer signals in real time during the data transmission process, predicts each communication anomaly or fault, and performs sandbox isolation testing on the potential fault interface; and the conflict resolution module dynamically eliminates multi-path data conflicts and optimizes the scheduling routing structure according to the detection results of each virtual channel.

[0035] Specifically, each superframe data packet in each virtual channel is converted by digital-quantum state encoding, the content of each superframe data packet is encoded into a corresponding quantum bit sequence, and then a microwave resonant cavity is used to establish a multi-node quantum entanglement channel between each virtual channel at the interface layer. For all virtual channels between each group, a path Hamiltonian model is constructed, and then each path Hamiltonian model is converted into a corresponding QUBO model. At the same time, all virtual channel combinations are initialized as uniformly distributed quantum superposition states, the total evolution time is set, and it is divided into multiple equally spaced time steps. The time evolution operator at each time step is updated in real time by using the Schrödinger equation to obtain the quantum state change rule of each group at each time step. Then, the quantum state of each group is updated step by step using the time evolution operator. During the evolution process at each time step, the tunneling strength is updated in real time based on the tunneling strength control mechanism, and the total Hamiltonian of the corresponding time step is adjusted according to the corresponding point tunneling strength. When the total evolution time is reached, the evolution is stopped, the current quantum state is measured, the probability amplitude of each virtual channel combination is obtained, and the virtual channel combination with the highest probability amplitude is taken as the optimal solution of the virtual channel combination between the corresponding groups. Then, the virtual channel combination is mapped to the virtual channel routing table, and the bandwidth resources are configured according to the node relationship.

[0036] Specifically, the specific calculation formula of digital-quantum state encoding conversion is as follows: In the formula, represents the mapped quantum bit; represents the content of the superframe data packet in the original data stream; represent complex probability amplitudes, and represent the standard ground state in quantum computing, representing logical 0 and 1; The specific form of the quantum superposition state is as follows: In the formula, represents the initial quantum superposition state; represents the total number of selectable virtual channels; represents each virtual channel combination; representing the quantum ground state corresponding to each virtual channel combination.

[0037] The security control module detects abnormal data transmitted by each virtual channel in real time, and traps and isolates each identified abnormal data.

[0038] Specifically, the transmission data of each interface is monitored in real time, and flow rate, connection frequency and source IP distribution are extracted from the monitored transmission data. At the same time, the flow entropy degree of the current network connection state is calculated according to the extracted flow characteristics. If the flow entropy degree of the adjacent time decreases by more than the preset abnormal threshold, it is judged as the DDoS behavior of botnet. Based on historical protection experience, manual setting and expert knowledge, multiple sets of abnormal behavior characteristics are set to establish corresponding antigen characteristic fingerprints. Then, based on each antigen characteristic fingerprint, multiple candidate antibodies are initialized to form an immune population. The recognition matching degree between each candidate antibody in the immune population and the antigen characteristic fingerprint is calculated. Each candidate antibody with a recognition matching degree higher than a preset threshold is selected, and the selected candidate antibodies are repeatedly cloned, mutated and screened until the recognition matching degree change value of each candidate antibody in the immune population converges within a preset termination threshold. Based on the final immune population, a corresponding response strategy model is established. The response strategy model analyzes whether each real-time transmission data is an attack behavior in real time. If it is judged as an attack behavior, the virtual bandwidth bait is activated, and the attackable target is simulated through the bandwidth bait to attract malicious data flow, while the occupied bandwidth used for attack in the real interface is released. When the response strategy model detects a new type of attack or antibody recognition fails, the corresponding transmission data is automatically added to the library as a new antigen sample, and the immune population of the response strategy model is updated.

[0039] The feedback regulation module is used for periodically analyzing system running data, entropy state trend and channel load, and optimizing the next round of scheduling strategy based on the analysis results.

Claims

1. A real-time data interface scheduling method for a multi-unit thermal system, characterized in that, The specific steps of this scheduling method are as follows: Ⅰ. Collect and process the operating data of each unit in real time, and analyze the local entropy change rate and trend of each unit based on the processed operating data; II. Based on the local entropy change rate and trend of each unit, dynamically adjust the scheduling priority of different data streams; III. Based on the scheduling priority of each data stream, construct a cross-unit virtual data channel, analyze the failure probability of each interface in real time, and conduct virtual isolation tests; IV. Based on the test results, dynamically adjust the bandwidth allocation of each interface, synchronously transmit each data stream command between multiple units, and resolve scheduling path conflicts in real time. V. Real-time detection of the total entropy of the multi-unit thermal system and dynamic reconstruction of the data path structure between interfaces; at the same time, real-time identification of abnormal data streams during data transmission. VI. After the scheduling action is completed, collect the performance indicators of this round of scheduling and optimize the scheduling strategy for the next round.

2. The real-time data interface scheduling method for a multi-unit thermal system according to claim 1, characterized in that, The specific steps for real-time acquisition and processing of the operating data of each unit described in Step I are as follows: S1.1: The multi-unit thermal system collects and classifies the operating data of each unit in real time through various sensors, and uses the moving average method to eliminate high-frequency noise in the various operating data. Then, the median of the smoothed operating data is selected as the dynamic benchmark value. S1.2: Based on each dynamic benchmark value, calculate the deviation rate of the corresponding types of operating data relative to the median. Then, based on the fluctuation characteristics of historical data, dynamically set the dynamic deviation threshold of each data transmission channel through the sliding standard deviation. If the deviation rate is higher than the current dynamic deviation threshold, the corresponding operating data will be marked as an outlier; otherwise, it will be marked as a normal value. S1.3: Count the number and location of missing and outlier values ​​in various types of operational data, and correct the counted missing and outlier values ​​by means of neighborhood interpolation. Then, standardize the corrected types of operational data to unify the dimensions of various types of operational data. At the same time, according to preset rules, unify the unit format of various types of operational data of each unit, and use Min-Max normalization to unify various types of operational data to the range of [0, 1].

3. The real-time data interface scheduling method for a multi-unit thermal system according to claim 2, characterized in that, The specific steps for analyzing the local entropy change rate and trend of each unit in Step I are as follows: S2.1: Extract the time series of each variable of each unit in the current thermodynamic cycle from the processed operating data. Based on the extracted time series of each variable of each unit, calculate the local entropy change rate of the multi-unit thermodynamic system in different time periods of the current thermodynamic cycle. S2.2: Smooth the calculated local entropy change rates of each group to eliminate the interference of instantaneous disturbances on the corresponding local entropy change rates. Then, based on the smoothed local entropy change rates, construct the state entropy value model of the multi-unit thermal system, and obtain the global state entropy value of the multi-unit thermal system in real time through the state entropy value model. S2.3: Real-time monitoring of the entropy value of the multi-unit thermal system at different times, and calculation of the change in the entropy value of the multi-unit thermal system at the current time relative to the previous time, so as to obtain the rate of change of the entropy value trend of the multi-unit thermal system. At the same time, based on historical experience and simulation results, the critical entropy value is set. S2.4: If the current entropy trend change rate is higher than the critical entropy value, it indicates that the current multi-unit thermal system is in a state of operational chaos and has a risk of control failure. Based on the local entropy change rate and global state entropy value of the current multi-unit thermal system, the corresponding scheduling feedback factor of each unit in the current multi-unit thermal system is calculated to obtain the current impact of each unit on the entropy increase of the multi-unit thermal system.

4. The real-time data interface scheduling method for a multi-unit thermal system according to claim 3, characterized in that, The specific calculation formula for the local entropy change rate mentioned in S2.1 is as follows: In the formula, Representing the The unit at time The rate of change of local entropy; Representing the The unit at time The heat output; Representing the The unit at time Heat input; Representing the The unit at time The effective thermal temperature; The specific calculation formula for the smoothing filter described in S2.2 is as follows: In the formula, Representing the The unit at time The smoothed rate of local entropy change; Represents the smoothing coefficient; Representing the The unit at time The smoothed rate of local entropy change; The specific form of the state entropy model described in S2.2 is as follows: In the formula, Represents a multi-unit thermal system at any time The global state entropy value; Represents the total number of generating units participating in the scheduling; Representing the Weighting factors for individual generating units; The specific formula for calculating the rate of change of entropy value trend mentioned in S2.3 is as follows: In the formula, represents the change in the global state entropy of the multi-unit thermal system at time t relative to the previous time t. Represents a multi-unit thermal system at any time The global state entropy value; The specific calculation formula for the scheduling feedback factor mentioned in S2.4 is as follows: In the formula, Representing the The scheduling feedback factor of each unit to the system entropy change; This represents a small constant used to avoid division by zero.

5. The real-time data interface scheduling method for a multi-unit thermal system according to claim 3, characterized in that, The specific steps for dynamically adjusting the scheduling priorities of different data streams as described in step II are as follows: S3.1: Based on the normalized operating data, establish data flow sets for each unit and its sub-modules, and mark each data flow set according to the preset unit number. Then, construct an "entropy value-data flow" correlation matrix based on the data flow set corresponding to each unit and the current smoothed local entropy change rate to obtain the entropy correlation weight between each unit and each data flow set in real time. S3.2: Perform full aggregation on the correlation weights of each entropy to obtain the correlation strength between each data stream set and the current multi-unit thermal system entropy. Then, normalize the obtained correlation strength of each data stream set and monitor the rate of change of the current entropy value in real time. S3.3: If the current entropy value trend change rate is lower than the critical entropy value, then based on the normalized correlation strength of each data stream set and the basic business weight allocated to the corresponding data stream, calculate the current scheduling priority factor of each data stream, and then arrange them in descending order according to the scheduling priority factor to form a scheduling queue. After the arrangement is completed, take out the data streams from the scheduling queue in turn to perform bandwidth allocation and transmission scheduling. S3.4: If the current entropy value trend change rate is higher than the critical entropy value, then based on the normalized correlation strength of each data stream set and the scheduling feedback factor, calculate the current scheduling priority factor of each data stream, and then arrange them in descending order according to the scheduling priority factor to form a scheduling queue. After the arrangement is completed, take out the data streams from the scheduling queue in turn to perform bandwidth allocation and transmission scheduling. S3.5: Based on the preset update time period, periodically re-extract the latest entropy change rate, check for any new or invalid data streams, and dynamically update the scheduling priority factor of each data stream according to the latest entropy change rate and the check results.

6. The real-time data interface scheduling method for a multi-unit thermal system according to claim 5, characterized in that, The specific steps for constructing cross-unit virtual data channels based on the scheduling priority of each data stream, as described in step III, are as follows: S4.1: Based on the current data stream scheduling priority ranking result, extract the data stream set in descending order of priority, record the corresponding source unit and target unit information, and establish a data path mapping table. S4.2: Collect the sampling timestamps of each data stream and the location information of the corresponding source unit in the logical topology of the current multi-unit thermal system, and add the collected sampling timestamps and source unit location information to the corresponding data stream set; S4.3: Calculate the spatiotemporal similarity between each data stream set. If the spatiotemporal similarity is higher than the preset similarity threshold, merge the corresponding data stream sets to construct a superframe data packet. Then, the multi-unit thermal system registers virtual channels for each superframe data packet at the interface layer. S4.4: After registration, based on the shortest latency strategy, find the physical path that meets the transmission constraints in the interface mapping matrix, and based on the search results, bind the virtual channel to the physical path obtained by the search, allocate a transmission time slot for each virtual channel and start channel transmission. S4.5: Based on the preset channel update cycle, the spatiotemporal similarity between each data stream set and the logical bandwidth requirements of each virtual channel are updated periodically. If the spatiotemporal correlation decreases or the bandwidth requirements exceed the preset threshold, the mapping relationship is automatically cancelled after the old virtual channel data is transmitted, physical resources are released, and a new virtual channel is dynamically constructed.

7. A real-time data interface scheduling system for a multi-unit thermal system, used to implement the real-time data interface scheduling method for a multi-unit thermal system according to any one of claims 1-6, characterized in that, It includes a data acquisition and processing module, an entropy change analysis module, a state assessment module, a dynamic priority module, a channel construction module, a prediction and simulation module, a conflict resolution module, a security control module, and a feedback regulation module; The data acquisition and processing module is used to acquire and process real-time operating data at high frequency from the data interfaces of each thermal power unit; The entropy change analysis module calculates the local entropy change rate of each unit based on the processed operating data. The status assessment module is used to integrate the local entropy change rate of each unit and to assess the current operating status of the multi-unit thermal system in real time based on the integration results. The dynamic priority module is used to generate scheduling priority factors for each data stream based on the real-time entropy weight distribution of the multi-unit thermal system, and to dynamically adjust the scheduling order of each data stream. The channel construction module is used to construct a reconfigurable virtual transmission channel at the interface layer and transmit data sequentially according to the scheduling order of each data stream; During data transmission, the prediction simulation module monitors the interface layer signals in real time, predicts various communication anomalies or faults, and performs sandbox isolation tests on potentially faulty interfaces. The conflict resolution module dynamically eliminates multi-path data conflicts and optimizes the scheduling routing structure based on the detection results of each virtual channel. The security control module detects abnormal data transmitted through each virtual channel in real time and traps and isolates the identified abnormal data. The feedback control module is used to periodically analyze system operation data, entropy trends, and channel load, and optimize the next round of scheduling strategy based on the analysis results.

8. A real-time data interface scheduling system for a multi-unit thermal system according to claim 7, characterized in that, The specific steps of the conflict resolution module in dynamically eliminating multi-path data conflicts and optimizing the scheduling routing structure are as follows: S5.1: The superframe data packets transmitted in real time in each virtual channel are encoded into corresponding quantum bit sequences through digital-quantum state encoding conversion. Then, a multi-node quantum entangled channel is established between each virtual channel in the interface layer using a microwave resonant cavity. S5.2: For all virtual channels between units, construct path Hamiltonian models respectively, then convert each path Hamiltonian model into a corresponding QUBO model, and initialize all virtual channel combinations as uniformly distributed quantum superposition states. S5.3: Set the total evolution time and divide it into multiple equally spaced time steps. Use the Schrödinger equation to update the time evolution operator at each time step in real time to obtain the quantum state change law of each unit at each time step. Then use the time evolution operator to gradually update the quantum state of each unit. S5.4: During the evolution process at each time step, based on the tunneling intensity control mechanism, the tunneling intensity at each time step is updated in real time, and the total Hamiltonian of the corresponding time step is adjusted according to the tunneling intensity at the corresponding point. When the total evolution time is reached, the evolution stops. S5.5: Measure the current quantum state, obtain the probability amplitude of each virtual channel combination, and take the virtual channel combination with the highest probability amplitude as the optimal solution for the virtual channel combination between the corresponding units. Then, map the virtual channel combination to the virtual channel routing table and configure bandwidth resources according to the node relationship.

9. A real-time data interface scheduling system for a multi-unit thermal system according to claim 7, characterized in that, The specific calculation formula for the digital-quantum state encoding conversion described in S5.1 is as follows: In the formula, Represents the mapped first One quantum bit; ; Represents the first in the original data stream The contents of a superframe data packet; as well as They represent the complex probability amplitudes, and These represent the standard ground state in quantum computing, indicating logic 0 and 1 respectively; The specific manifestations of the quantum superposition state described in S5.2 are as follows: In the formula, Represents the initial quantum superposition state; Represents the total number of available virtual channels; Represents each virtual channel combination; This represents the quantum ground state corresponding to each virtual channel combination.

10. A real-time data interface scheduling system for a multi-unit thermal system according to claim 7, characterized in that, The specific steps by which the security control module traps and isolates the identified abnormal data are as follows: S6.1: Monitor the transmission data of each interface in real time, and extract the flow rate, connection frequency and source IP distribution of each traffic feature from the monitored transmission data. At the same time, calculate the traffic entropy of the current network connection state based on the extracted traffic features. If the decrease in traffic entropy at adjacent times exceeds the preset abnormal threshold, it is judged as DDoS behavior of botnet. S6.2: Based on historical protection experience, manual settings and expert knowledge, set multiple sets of abnormal behavior features to establish corresponding antigen feature fingerprints. Then, based on each antigen feature fingerprint, initialize multiple sets of candidate antibodies to form an immune population. Then calculate the recognition matching degree between each candidate antibody in the immune population and the antigen feature fingerprint. S6.3: Select each candidate antibody whose identification matching degree is higher than the preset threshold, and repeat the cloning, mutation and screening operations for each candidate antibody until the identification matching degree change value of each candidate antibody in the immune population converges to within the preset termination threshold, and establish a corresponding response strategy model based on the final immune population. S6.4: The response strategy model analyzes in real time whether each real-time transmitted data is an attack. If it is determined to be an attack, it will activate the virtual bandwidth decoy and use the bandwidth decoy to simulate an attackable target to attract malicious data streams. At the same time, it will release the bandwidth occupied by the real interface for the attack. When the response strategy model detects a new attack or antibody recognition failure, it will automatically add the corresponding transmitted data as a new antigen sample to the library and update the immune population of the response strategy model.

Images