Low-voltage power supply redundancy switching control method applied to air separation system

Abstract

The present invention relates to the field of power control technology, and in particular to a low-voltage power supply redundancy switching control method applied to an air separation system, including a method applied to an air separation system comprising at least two independent low-voltage power supply input lines, and configuring an emergency backup power supply as a third redundant input line, comprising: real-time acquisition of a power supply status data set of a low-voltage power distribution system; extraction of available features on the power supply status data set to obtain available characterization parameters of each input line; in response to any item in the available characterization parameters not meeting a corresponding preset input line availability threshold, determining that the input line is unavailable; otherwise, marking it as available; calculating power supply redundancy based on the number of available input lines and the total number of input lines; in response to the power supply redundancy being lower than a preset redundancy threshold, obtaining a switching decision feature set; inputting the switching decision feature set into a power supply redundancy control model to obtain an optimal switching parameter set, and performing low-voltage power supply redundancy switching control based on the optimal switching parameter set.

Description

Technical Field

[0001] The present invention relates to the technical field of power control, and in particular to a low-voltage power supply redundancy switching control method applied to an air separation system. Background Art

[0002] As the core equipment for industrial gas production, the stability of the low-voltage power distribution system of an air separation system directly determines the safe operation and production efficiency of the entire system. As the energy foundation for driving circuit breakers, control systems, and critical loads, the reliability of the low-voltage power supply is crucial.

[0003] In the existing technology, low-voltage power supply systems generally adopt a dual-circuit redundant architecture to achieve basic power supply security, but there are still multiple failure risks under complex working conditions; traditional redundant switching control mainly relies on electrical interlocking protection and manual intervention to start the emergency power supply, and cannot adaptively optimize switching parameters. It has poor adaptability in complex and changeable abnormal scenarios; at the same time, the operation mode that relies on manual intervention to start the emergency power supply has a response lag, which can easily cause system downtime or equipment damage; there is a lack of dynamic quantification of the coordination capability of multiple power supplies, and it is difficult to accurately trigger the switching logic. Not only can the advantages of power redundancy configuration not be fully utilized, but it may also cause power supply fluctuations during the switching process. In severe cases, it may even cause a brief power outage in the air separation system, affecting the stable operation of the system. Summary of the Invention

[0004] The present invention provides a low-voltage power supply redundancy switching control method applied to an air separation system, which improves the stability of the low-voltage power distribution system of the air separation system, reduces system shutdown accidents caused by power supply problems, and thus improves production efficiency. The method can effectively solve the problems in the background technology.

[0005] To achieve the above-mentioned object, the present invention provides a low-voltage power supply redundancy switching control method for an air separation system. The method is applied to an air separation system comprising at least two independent low-voltage power supply input lines, and an emergency backup power supply is configured as a third redundant input line, comprising:

[0006] Real-time acquisition of power status data sets of low-voltage distribution systems;

[0007] Extracting available features from the power status data set to obtain available characterization parameters for each input line;

[0008] In response to any one of the available characterization parameters not meeting the corresponding preset input line availability threshold, determining that the input line is unavailable; otherwise, marking it as available;

[0009] The power redundancy is calculated based on the number of available input lines and the total number of input lines;

[0010] In response to the power supply redundancy being lower than a preset redundancy threshold, obtaining a switching decision feature set;

[0011] The switching decision feature set is input into a power supply redundancy control model to obtain an optimal switching parameter set, and low-voltage power supply redundancy switching control is performed based on the optimal switching parameter set.

[0012] In combination with the first aspect, in one possible design, the power status data set includes a low-voltage side data subset of each input line, an emergency backup power supply data subset, and a circuit breaker data subset;

[0013] The low-voltage side data subset includes the low-voltage side output voltage, load rate and historical power failure frequency of each incoming cabinet;

[0014] The emergency backup power supply data subset includes voltage stability characteristics, readiness status and available time;

[0015] The circuit breaker data subset includes coil temperature, closing status and duration of instantaneous voltage loss of control power supply.

[0016] In combination with the first aspect, in a possible design, the available characterization parameters are any one or a combination of two or more of voltage deviation, load rate, number of consecutive power failures, continuous power supply time of a single power supply, and ambient temperature deviation.

[0017] In combination with the first aspect, in one possible design, the switching decision feature set includes power failure segment, power failure type, load priority, and backup power supply matching degree.

[0018] In combination with the first aspect, in one possible design, the optimal switching parameter set includes a switching path, a switching delay, and load classification control.

[0019] In combination with the first aspect, in a possible design, the preset input line threshold value can be set based on voltage deviation, load rate, number of consecutive power failures, continuous power supply time of a single power supply, and ambient temperature deviation.

[0020] In combination with the first aspect, in a possible design, the calculation formula for power redundancy is: Where R represents power redundancy, Na represents the number of available input lines, and Nt represents the total number of input lines.

[0021] In conjunction with the first aspect, in one possible design, the method for acquiring the handover decision feature set includes:

[0022] By monitoring the power status dataset of the low-voltage distribution system in real time, the line section that has lost power is determined and located, and the monitored power outage information is integrated into power outage section feature information;

[0023] Combined with the historical data and real-time changes of the output voltage on the low-voltage side of the input line, as well as the changes in the load factor, the type of power failure is analyzed and obtained;

[0024] Evaluate and classify each load based on the air separation system's operating requirements and the load's importance to obtain load priority information;

[0025] Based on the emergency backup power data subset, evaluate whether the current status of the backup power supply meets the switching requirements; based on the backup power supply status information and system requirements, calculate the matching information between the backup power supply and the current power-off line section;

[0026] The power failure segment feature information, power failure type information, load priority information and backup power supply matching information are integrated to obtain the switching decision feature set.

[0027] In conjunction with the first aspect, in one possible design, the structure of the power redundancy control model includes:

[0028] An input layer, used to receive a switching decision feature set as input;

[0029] Hidden layer, including at least one neural network layer, used to perform nonlinear transformation and feature extraction on input features;

[0030] Output layer, used to output the optimal switching parameter set;

[0031] The control logic layer is used to generate corresponding control instructions based on the optimal switching parameter set of the output layer.

[0032] In combination with the first aspect, in one possible design, the power redundancy control model uses a deep learning algorithm as the basic structure of the model.

[0033] The technical solution of the present invention can achieve the following technical effects:

[0034] The method forms a comprehensive system risk monitoring system by real-time monitoring of the power status dataset of the low-voltage distribution system and combining available feature extraction and redundancy evaluation; it can detect potential multiple failure risks in advance, rather than just the failure of a single component, thereby significantly reducing the overall failure probability of the entire low-voltage distribution system due to complex working conditions; traditional redundant switching control relies on electrical interlocking protection and manual intervention, while this method realizes the intelligent and adaptive optimization of switching decisions by introducing a power redundancy control model; in complex and changeable abnormal scenarios, the system can automatically adjust the switching parameters according to real-time data without manual intervention, greatly improving the adaptability and response speed of the system and reducing switching failures or response lags caused by human factors; the method calculates power redundancy and triggers switching decisions when the redundancy is lower than the preset threshold, ensuring that the system can maintain sufficient power redundancy at any time; combined with the optimal switching parameter set, it can minimize power supply fluctuations during the switching process, The risk of causing a brief power outage to the air separation system is avoided, thereby significantly enhancing the stability of power supply; by dynamically quantifying the collaborative capabilities of multiple power supplies and accurately triggering the switching logic, this method can fully utilize the advantages of power redundancy configuration; when switching is required, the system can select the optimal switching path and load classification control strategy to ensure continuous power supply to critical loads, while reducing the pressure of non-critical loads on the power system, thereby maximizing the benefits of power redundancy configuration; the automation and intelligent characteristics of the method reduce the need for manual intervention and reduce operational complexity; at the same time, through real-time monitoring and early warning mechanisms, potential problems can be discovered and handled in advance, reducing downtime and maintenance costs caused by system failures; by improving the stability and power supply reliability of the low-voltage distribution system, this method indirectly improves the overall production efficiency of the air separation system; reduces system shutdown accidents caused by power problems, ensures the continuity and stability of the industrial gas preparation process, and thus improves production efficiency and product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 is a flow chart of the present invention;

[0036] Figure 2 This is a structural diagram of the power redundancy control model in the present invention. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0038] The present application is described below in conjunction with the accompanying drawings.

[0039] like Figure 1As shown, the low-voltage power supply redundancy switching control method applied to an air separation system of the present invention specifically includes the following steps:

[0040] S1. Real-time acquisition of power status data set of low-voltage power distribution system;

[0041] S2. Extract available features from the power status data set to obtain available characterization parameters for each input line;

[0042] S3. In response to any one of the available characterization parameters not meeting the corresponding preset input line availability threshold, determining that the input line is unavailable; otherwise, marking it as available;

[0043] S4. Calculate power redundancy based on the number of available input lines and the total number of input lines;

[0044] S5. In response to the power supply redundancy being lower than a preset redundancy threshold, obtaining a switching decision feature set;

[0045] S6. Input the switching decision feature set into a power supply redundancy control model to obtain an optimal switching parameter set, and perform low-voltage power supply redundancy switching control based on the optimal switching parameter set.

[0046] In this embodiment, the method obtains the power status dataset of the low-voltage power distribution system in real time, including the low-voltage side data subset of each input line, the emergency backup power data subset, and the circuit breaker data subset, ensuring comprehensive monitoring of the system status; extracts usable features from the power status dataset to obtain usable characterization parameters for each input line, thereby improving the accuracy and reliability of the judgment;

[0047] The power redundancy is calculated based on the number of available input lines and the total number of input lines. This method can adaptively assess the redundancy level of the current system and provide a basis for subsequent switching decisions. When the power redundancy falls below a preset redundancy threshold, the method obtains a switching decision feature set and, through a power redundancy control model, obtains the optimal switching parameter set. This achieves intelligent and automated switching decisions, reduces the need for manual intervention, and improves response speed and switching accuracy.

[0048] The optimal switching parameter set includes switching paths, switching delays, and load classification control, which helps reduce power supply fluctuations during the switching process, ensures stable system operation, and avoids system downtime or equipment damage caused by improper switching.

[0049] This method can significantly improve the stability of the low-voltage power distribution system of the air separation system, reduce system shutdown accidents caused by power problems, and thus improve production efficiency. Through real-time monitoring and intelligent decision-making, the method can prevent multiple failure risks under complex working conditions and ensure the reliable operation of the power supply system.

[0050] In summary, this method effectively solves the problems of multiple failure risks, delayed response to manual intervention, and power supply fluctuations in the low-voltage distribution system of the air separation system through real-time data monitoring, precise feature extraction, adaptive redundancy assessment, intelligent switching decision-making, and optimized switching parameter control, significantly improving the system stability and production efficiency.

[0051] In some embodiments of the present invention, for step S1:

[0052] The power status data set includes a low-voltage side data subset of each input line, an emergency backup power supply data subset, and a circuit breaker data subset;

[0053] The low-voltage side data subset includes the low-voltage side output voltage, load factor, and historical power outage frequency of each incoming cabinet. Monitoring low-voltage side voltage fluctuations directly reflects the quality of the main power supply and is a key indicator for determining whether the power supply is normal. The load factor reflects the actual load capacity of the power supply. When it exceeds 80%, an overload risk warning is required to avoid insulation aging or over-tripping. The historical power outage frequency is used to identify power supply stability trends. High-frequency power outages may indicate poor line contact or equipment aging.

[0054] The emergency backup power supply data subset includes voltage stability characteristics, readiness status, and available time. Dynamic indicators such as the ripple coefficient and harmonic content of the backup power supply output voltage are monitored to ensure that the power quality after switching meets the requirements of precision control equipment. The readiness status includes the start and stop status of the backup power supply and the charge state of the energy storage device. The available time quantifies the sustainable power supply time of the backup power supply and is used for redundancy calculation and switching strategy formulation.

[0055] The circuit breaker data subset includes coil temperature, closing status and duration of instantaneous power loss of control power supply; monitoring the circuit breaker operating coil temperature. Overtemperature may indicate coil insulation degradation or mechanical jamming, resulting in failure to open and close the circuit breaker; real-time confirmation of the position of the circuit breaker main contacts to avoid control logic confusion caused by false alarms of auxiliary contacts; and recording of short-term power loss events in the control circuit. Such faults are easily ignored by traditional protection devices, but may cause the circuit breaker to trip incorrectly or the control system to freeze.

[0056] In this embodiment, by collecting the output voltage, load rate, and historical power failure frequency of the low-voltage side, the power supply quality, load health status, and stability trend of the main power supply can be judged in real time, and hidden problems such as poor line contact and equipment aging can be discovered in advance, avoiding missed faults caused by misjudgment of a single indicator; the voltage stability characteristics, ready status, and available time of the backup power supply are monitored to ensure that the backup power supply has qualified power quality and continuous power supply capability when switching, avoiding switching failures caused by the unavailability of the backup power supply; collecting the coil temperature, closing status, and instantaneous power loss duration of the control power supply can timely discover microsecond-level faults that are difficult to capture with traditional protection, such as mechanical jamming of the circuit breaker, false alarm of the contact, and instantaneous disconnection of the control circuit, to prevent system loss of control due to abnormal actuators; through industrial-grade sensors The device and real-time communication network achieve millisecond-level data collection and transmission, ensuring the timeliness of subsequent feature extraction, redundancy calculation and switching decisions, and meeting the strict millisecond-level response requirements of the air separation system to power outages; different subsets of data can verify each other, improving the accuracy of risk judgment; based on data such as historical power outage frequency and coil temperature trends, power supply equipment and circuit breakers can be inspected and maintained to reduce the probability of unplanned downtime; the switching priority can be dynamically adjusted based on the standby power readiness status and available time data to ensure a fast and smooth switching process and reduce the impact of power supply fluctuations on the air separation system; a universal protocol is used to collect multi-type sensor data, which is compatible with equipment of different brands, facilitating system expansion to complex architectures with multiple power supplies and multiple loads, and reducing subsequent upgrade costs.

[0057] In some embodiments of the present invention, for step S2:

[0058] The available characterization parameters include any one or a combination of two or more of voltage deviation, load rate, number of consecutive power failures, continuous power supply time of a single power supply, and ambient temperature deviation;

[0059] Voltage deviation reflects the difference between the power supply output voltage and the rated voltage and is an important indicator for evaluating power supply quality. Excessive voltage deviation may cause device damage or performance degradation.

[0060] The load factor indicates the ratio of the current load to the rated capacity of the line and is a key parameter for evaluating line load conditions. An excessively high load factor may trigger overload protection, leading to line interruption.

[0061] The number of consecutive power outages records the number of consecutive power outages in a short period of time and is an important indicator for evaluating line stability and reliability. Frequent consecutive power outages may indicate serious problems in the line.

[0062] The continuous power supply duration of a single power source reflects the continuous power supply capability of the line without backup power support;

[0063] Although ambient temperature deviation does not directly reflect the power supply status, changes in ambient temperature may affect the heat dissipation and insulation performance of the equipment, thereby indirectly affecting the availability of the power supply.

[0064] Clean, denoise and normalize the raw data to improve data quality and consistency;

[0065] Calculate the characteristic value corresponding to each input line based on the selected available characterization parameters; voltage deviation can be calculated by comparing the difference between the actual output voltage and the rated voltage; load factor can be calculated by the ratio of the current load current to the line rated current;

[0066] Based on actual needs, multiple features can be combined to form a more comprehensive availability evaluation index; voltage deviation and load rate can be combined to jointly evaluate the power quality and load conditions of the line;

[0067] By extracting available features from the power status dataset, we can obtain available characterization parameters for each input line. These parameters can accurately reflect the line availability status, providing strong support for subsequent line availability judgment and power redundancy calculation. Feature extraction also helps reduce the complexity and computational effort of data processing, improving the system's response speed and decision-making efficiency.

[0068] In this embodiment, by extracting core parameters such as voltage deviation and load rate, the power supply output quality and load level can be accurately quantified, avoiding misjudgments caused by vague or one-sided assessments. The number of consecutive power outages and the duration of continuous power supply from a single power source constitute a dynamic early warning system. The former can identify poor line contact or aging trends, while the latter can assess the critical risk duration before redundant power supply switching. Combined with the indirect impact of ambient temperature deviation on insulation aging, a multi-dimensional fault prediction model can be established to achieve hierarchical prevention and control from the device level to the system level. Data cleaning and normalization eliminate noise interference and dimensional differences, making different features comparable and ensuring the convergence and stability of subsequent model training. The composite index formed by combining multiple features can construct a more accurate availability evaluation function, achieving an upgrade from static redundancy configuration to dynamic demand matching, avoiding resource waste caused by excessive redundancy or system crashes caused by insufficient redundancy. Feature extraction compresses the original data dimensions into a set of key parameters, reducing the computational complexity of subsequent availability judgment and redundancy calculation, significantly shortening the decision cycle. The modular feature extraction framework supports the dynamic expansion of new parameters and achieves continuous adaptation of the system to changing operating conditions through a weight adaptive adjustment algorithm.

[0069] In some embodiments of the present invention, for step S3:

[0070] The basis for setting the threshold value of the preset input line includes:

[0071] Voltage deviation: Set an allowable voltage deviation range based on the device's voltage stability requirements. When the actual voltage deviation exceeds this range, it will affect the normal operation of the device or cause performance degradation.

[0072] Load rate: Set an upper limit for the load rate based on the rated capacity of the line and the actual load requirements of the equipment. When the load rate exceeds 80%, it may trigger overload protection or cause severe heating of the line, affecting the stability and safety of the system.

[0073] The number of consecutive power outages: Based on historical data and experience, set an upper limit for the number of consecutive power outages allowed. Frequent consecutive power outages may indicate serious line problems, such as poor contact or aging equipment, which require timely treatment.

[0074] The duration of continuous power supply by a single power source is determined by setting a lower limit based on the system's redundancy configuration and the startup time of the emergency power source. If the duration of continuous power supply by a single power source falls below this lower limit, it may mean that the system is about to lose redundancy and the emergency power source needs to be activated as soon as possible or other measures need to be taken.

[0075] Ambient temperature deviation: Although ambient temperature deviation does not directly reflect the power supply status, changes in ambient temperature may affect the heat dissipation and insulation performance of the device. Therefore, it is necessary to set an allowable ambient temperature deviation range to ensure that the device operates in an appropriate temperature environment.

[0076] For each input line, the extracted available characterization parameters are compared with the corresponding preset available thresholds one by one;

[0077] If any available characterization parameter does not meet the corresponding preset availability threshold, the input line is determined to be unavailable; conversely, if all available characterization parameters meet the corresponding preset availability threshold, the input line is marked as available;

[0078] By determining the availability of each input line in real time, potential power supply problems or line failures can be discovered in a timely manner, providing early warning for stable system operation.

[0079] The availability judgment results provide an important basis for subsequent power redundancy calculations and switching decisions. When an input line is determined to be unavailable, it can be removed from the list of available lines, thus affecting the power redundancy calculation results.

[0080] After determining that a certain input line is unavailable, appropriate measures can be taken to isolate or repair the fault to prevent the fault from expanding or affecting the normal operation of other lines.

[0081] In this embodiment, by comparing the available characterization parameters of each input line with the preset thresholds one by one, it is possible to capture abnormal conditions such as voltage deviation, load rate exceeding the limit, and continuous power failure in real time, and achieve millisecond-level warning of faults, thereby avoiding performance degradation or damage of equipment due to voltage abnormalities; the availability judgment result directly affects the power redundancy calculation; when a line is judged to be unavailable, the system can automatically remove it from the list of available lines and recalculate the redundancy of the remaining lines; ensure the dynamic adaptability of the redundancy strategy and avoid the risks caused by insufficient redundancy; once a line is judged to be unavailable, the system can quickly trigger the fault isolation mechanism, such as cutting off the faulty line by opening the circuit breaker to prevent the fault from spreading to other healthy lines; at the same time, the system can record the mark of the unavailable line Identification provides precise positioning information for subsequent maintenance and shortens fault recovery time; real-time availability judgment provides a health status snapshot for the system, ensuring power supply continuity under complex working conditions; the availability judgment results provide basic data for the generation of subsequent switching decision feature sets; through automated availability judgment, the system can reduce the frequency of manual inspections and shorten the fault response time from hours to seconds; at the same time, historical unavailability records can provide data support for preventive maintenance and reduce long-term operation and maintenance costs; the dynamic availability judgment mechanism enables the system to have "self-diagnosis" capabilities and maintain high availability in complex and changeable industrial environments; when the continuous power supply duration of a single power supply is close to the critical value for starting the backup power supply, the system can preheat the backup power supply in advance to ensure that the switching process is imperceptible.

[0082] In some embodiments of the present invention, for step S4:

[0083] Power redundancy refers to the ratio of the number of available power lines to the total number of power lines in the system. It reflects the system's ability to cope with power failures. The higher the redundancy, the stronger the system's ability to maintain power supply when some power lines fail, thereby improving system reliability and stability.

[0084] Because power supply status may change over time, power redundancy calculations must be performed in real time to reflect the system's current power supply reliability. Accurate availability assessments are a prerequisite for calculating power redundancy. Any misjudgment can lead to inaccurate redundancy calculations, impacting subsequent switching decisions. During system operation, power lines may be added, removed, or adjusted based on actual needs. Therefore, power redundancy calculations must be able to dynamically adapt to these changes.

[0085] The calculation formula for power redundancy is: Where R represents power redundancy, Na represents the number of available input lines, and Nt represents the total number of input lines.

[0086] In this embodiment, by calculating the power redundancy, the abstract concept of system power supply reliability is converted into a specific value, thereby realizing a quantitative evaluation of the system power supply capacity; this helps the operation and maintenance personnel to intuitively understand the current power supply status of the system and provide data support for subsequent decision-making; given that the power supply status may change over time, this step emphasizes the real-time calculation of power redundancy to ensure that the system can continuously and dynamically reflect the power supply reliability; this helps to timely discover power supply hidden dangers and avoid the accumulation of potential risks; the accuracy of the power redundancy calculation results is directly related to the rationality of subsequent switching decisions; this step provides a reliable basis for switching decisions by ensuring the accuracy of availability judgment, thereby reducing switching failures or power supply failures caused by misjudgment. Fluctuation risk; the power redundancy calculation method has dynamic adaptability to the increase, decrease or adjustment of power lines that may occur during system operation; ensuring that redundancy can be accurately calculated regardless of how the system changes, providing continuous protection for the stable operation of the system; by calculating power redundancy in real time and accurately, the system can take timely measures when redundancy is insufficient, thereby avoiding or reducing system downtime or equipment damage caused by power failure, significantly improving system stability and security; the power redundancy calculation results can provide strong support for operation and maintenance management; operation and maintenance personnel can predict potential problems based on the redundancy change trend, formulate maintenance plans in advance, implement preventive maintenance, reduce operation and maintenance costs, and improve operation and maintenance efficiency.

[0087] In some embodiments of the present invention, for step S5:

[0088] The switching decision feature set is a collection of key information used to guide switching operations when the system faces insufficient power redundancy. The switching decision feature set includes power failure segment, power failure type, load priority, and backup power matching, which can fully reflect the current status and needs of the system.

[0089] The power outage segment specifies the specific line segment or equipment area where the power outage occurred. This helps operations and maintenance personnel quickly locate the fault point and assess the impact of the power outage on the entire system. In switching decisions, the power outage segment information can be used to determine which loads require priority for power restoration and which backup power sources or switching paths are available.

[0090] The power outage type describes the specific cause or nature of the power outage, such as momentary power outage, permanent power outage, or voltage fluctuation. Different power outage types require different switching strategies. A momentary power outage may require automatic recovery after a short wait, while a permanent power outage requires immediate activation of a backup power source or switching to another available line.

[0091] Load priority is the ranking of loads based on their importance to system operation. Critical loads have higher priority and their power supply must be prioritized during the handover process. Load priority information is used to guide load classification control during handover, ensuring that important loads are not affected or are minimally affected during the handover process.

[0092] Backup power supply matching is used to assess the degree of compatibility between backup power supplies and current system requirements, including aspects such as backup power supply capacity, voltage stability, and readiness. In switching decisions, backup power supply matching information can be used to select the most appropriate backup power supply for switching, thereby avoiding switching failures or power supply fluctuations caused by backup power mismatch.

[0093] The method for obtaining the switching decision feature set includes:

[0094] By real-time monitoring of the power status dataset of the low-voltage distribution system, the specific line section where the power outage occurred is located. The circuit breaker closing status and the duration of the instantaneous power outage are used to determine the line section that has lost power. The monitored power outage information is integrated into power outage segment feature information.

[0095] Combining the historical data and real-time changes of the output voltage on the low-voltage side of the input line, as well as the changes in the load rate, the system can analyze the specific type of power failure;

[0096] Evaluate and classify each load based on the air separation system's operating requirements and the load's importance to obtain load priority information;

[0097] Emergency backup power supply data subset, evaluates whether the current status of the backup power supply meets the switching requirements; calculates the matching information between the backup power supply and the current power-off line segment based on the backup power supply status information and system requirements;

[0098] The power failure section, power failure type, load priority, and backup power supply matching information obtained in the above steps are integrated to obtain a switching decision feature set.

[0099] In this embodiment, by real-time monitoring and integration of power outage segment information, the specific line segment or equipment area where the power outage occurred can be quickly located, helping operations and maintenance personnel to quickly respond and assess the impact of the power outage on the entire system, thereby providing an accurate basis for subsequent switching decisions. By analyzing the power outage type information, the system can adopt corresponding switching strategies based on the cause or nature of the power outage, ensuring the efficiency and accuracy of the switching process and avoiding unnecessary switching operations or response delays. Based on load priority information, the system can prioritize the power supply to critical loads during the switching process, ensuring the stable operation of important equipment and systems and reducing the risk of production interruption or equipment damage caused by switching. By evaluating the backup power supply matching information, the system can select the most appropriate backup power supply for switching, avoiding switching failures or power supply fluctuations caused by backup power mismatch, thereby improving the success rate of switching and the power supply reliability of the system. Integrating information such as power outage segment, power outage type, load priority, and backup power supply matching into a switching decision feature set provides comprehensive and accurate decision support for the power redundancy control model, helping the system make optimal switching decisions in complex and changing abnormal scenarios, thereby improving system stability and reliability and ensuring the safe operation and production efficiency of the air separation system.

[0100] In some embodiments of the present invention, for step S6, as shown in 2:

[0101] The optimal switching parameter set includes a switching path, a switching delay, and a load classification control;

[0102] The power redundancy control model determines the optimal switching path based on the power-loss section and backup power matching information in the switching decision feature set. This includes selecting which backup power source to switch to and how to implement the switching operation through circuit breakers and other power distribution equipment. The selection of the switching path takes into account factors such as the power source capacity, voltage stability, readiness, and availability time to ensure the quality of power after the switch.

[0103] The switching delay model dynamically adjusts the switching delay based on the type of power outage and load priority information. The adjustment of the switching delay aims to balance the relationship between switching speed and system stability, avoiding power supply fluctuations or equipment damage caused by switching too quickly or too slowly.

[0104] Load classification control: Based on load priority information, the model will formulate a load classification control strategy. During the switching process, critical loads will be given priority in power supply, while secondary loads may be temporarily powered off or load reduced depending on the situation. Load classification control helps ensure power supply continuity for important loads during the switching process, reducing the risk of production interruption or equipment damage caused by the switching.

[0105] The power redundancy control model uses a deep learning algorithm as the basic structure of the model. The structure of the power redundancy control model includes:

[0106] The input layer receives the switching decision feature set as input, including the power failure section, power failure type, load priority, and backup power matching degree.

[0107] Hidden layer, which consists of one or more neural network layers, is used to perform nonlinear transformation and feature extraction on the input features. Each layer is connected by weights and biases, which are adjusted by the back-propagation algorithm during training.

[0108] The output layer is used to output the optimal switching parameter set, including parameters such as switching path, switching delay, and load classification control. The output layer can be designed in different forms according to specific needs, such as classification output, regression output, etc.

[0109] The control logic layer is used to generate corresponding control instructions based on the optimal switching parameter set of the output layer; the control instructions are sent to the low-voltage power distribution system through the interface program to realize power supply redundancy switching control.

[0110] In this embodiment, through the power redundancy control model, the system can adaptively optimize the switching parameters according to the switching decision feature set obtained in real time; ensure that the system can make the most appropriate switching decision under different complex working conditions, and improve the flexibility and adaptability of the switching control; the model performs comprehensive analysis based on multi-dimensional features, and can accurately judge when power switching is required and determine the optimal switching path; avoid the limitations of traditional methods that rely on a single electrical interlocking protection or manual intervention, reduce the possibility of misjudgment and missed judgment, and improve the accuracy of the switching logic; by dynamically adjusting the switching delay time, the model can find a balance between switching speed and system stability; avoid power supply fluctuations caused by switching too fast or too slow, ensure the continuity and stability of power supply during the switching process, and reduce Impact on key equipment such as air separation systems; the implementation of load classification control strategy ensures that key loads can be given priority in power supply during the switching process; reduces the risk of production interruption or equipment damage due to switching, and improves the reliability and production efficiency of the system; the application of power redundancy control model gives full play to the advantages of power redundancy configuration; through precise triggering of switching logic and adaptive optimization of switching parameters, the system can make switching decisions quickly and accurately under complex working conditions, ensuring the continuity and stability of power supply, and improving the reliability and safety of the entire air separation system; this step realizes the intelligent switching control of low-voltage power redundancy by introducing power redundancy control model; this reduces the need for manual intervention, reduces operation complexity and response lag, and improves the automation level and operation efficiency of the system.

[0111] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A low-voltage power supply redundancy switching control method applied to an air separation system, characterized in that: The method is applied to an air separation system comprising at least two independent low-voltage power input lines, and an emergency backup power supply is configured as a third redundant input line, comprising: Real-time acquisition of power status data sets of low-voltage distribution systems; Extract available features from the power status data set to obtain available characterization parameters for each input line; the available characterization parameters are any one or a combination of two or more of voltage deviation, load rate, number of consecutive power outages, continuous power supply duration of a single power source, and ambient temperature deviation; In response to any one of the available characterization parameters not meeting the corresponding preset input line availability threshold, determining that the input line is unavailable; otherwise, marking it as available; The power redundancy is calculated based on the number of available input lines and the total number of input lines; In response to the power redundancy being lower than a preset redundancy threshold, obtaining a switching decision feature set; the switching decision feature set includes a power failure segment, a power failure type, a load priority, and a backup power supply matching degree; Inputting the switching decision feature set into a power supply redundancy control model to obtain an optimal switching parameter set, and performing low-voltage power supply redundancy switching control based on the optimal switching parameter set; The method for obtaining the switching decision feature set includes: By monitoring the power status dataset of the low-voltage distribution system in real time, the line section that has lost power is determined and located, and the monitored power outage information is integrated into power outage section feature information; Combined with the historical data and real-time changes of the output voltage on the low-voltage side of the input line, as well as the changes in the load factor, the type of power failure is analyzed and obtained; Evaluate and classify each load based on the air separation system's operating requirements and the load's importance to obtain load priority information; Based on the emergency backup power data subset, evaluate whether the current status of the backup power supply meets the switching requirements; based on the backup power supply status information and system requirements, calculate the matching information between the backup power supply and the current power-off line section; The power failure segment feature information, power failure type information, load priority information, and backup power supply matching information are integrated to obtain a switching decision feature set; The method for obtaining the available characterization parameters includes: The voltage deviation is calculated by comparing the difference between the actual output voltage and the rated voltage; The load factor is calculated by the ratio of the current load current to the line rated current; The number of consecutive power failures indicates the number of consecutive power failures within the set time. The single power supply continuous power supply duration refers to the continuous power supply duration of the line without backup power support.

2. The low-voltage power supply redundancy switching control method applied to an air separation system according to claim 1 is characterized in that: The power status data set includes a low-voltage side data subset of each input line, an emergency backup power supply data subset, and a circuit breaker data subset; The low-voltage side data subset includes the low-voltage side output voltage, load rate and historical power failure frequency of each incoming cabinet; The emergency backup power supply data subset includes voltage stability characteristics, readiness status and available time; The circuit breaker data subset includes coil temperature, closing status and duration of instantaneous voltage loss of control power supply.

3. The low-voltage power supply redundancy switching control method applied to an air separation system according to claim 1 is characterized in that: The optimal handover parameter set includes a handover path, a handover delay, and load classification control.

4. The low-voltage power supply redundancy switching control method applied to an air separation system according to claim 1 is characterized in that: The preset input line available threshold value may be set based on voltage deviation, load rate, number of consecutive power failures, duration of continuous power supply by a single power source, and ambient temperature deviation.

5. The low-voltage power supply redundancy switching control method applied to an air separation system according to claim 1 is characterized in that: The calculation formula for power redundancy is: ； Where R represents power redundancy, Na represents the number of available input lines, and Nt represents the total number of input lines.

6. The low-voltage power supply redundancy switching control method applied to an air separation system according to claim 1, characterized in that: The structure of the power redundancy control model includes: An input layer, used to receive a switching decision feature set as input; Hidden layer, including at least one neural network layer, used to perform nonlinear transformation and feature extraction on input features; Output layer, used to output the optimal switching parameter set; The control logic layer is used to generate corresponding control instructions based on the optimal switching parameter set of the output layer.

7. The low-voltage power supply redundancy switching control method applied to an air separation system according to claim 6, characterized in that: The power supply redundancy control model adopts a deep learning algorithm as the basic mechanism of the model.

Images