PLC control-based UPS intelligent control system for ship

Abstract

The application relates to the technical field of ship electrical automation, and specifically provides a marine UPS intelligent control system based on PLC control, which comprises an edge computing unit for extracting salt mist-vibration coupling stress characteristics; a battery state estimation module adopts an adaptive fusion filtering algorithm with coupling stress correction, a correction process noise covariance matrix and an internal resistance aging term are corrected, and the battery state of charge and the health state are jointly and on-line estimated; a PLC master control unit performs nonlinear multidimensional correction on a basic charging current, generates a charging current reference value and outputs a control instruction; an execution and feedback module feeds back actual execution parameters and battery operation states to the PLC master control unit and the edge computing unit respectively, so that a closed-loop optimization is formed. The application solves the problem of battery life prediction inaccuracy caused by unidentified salt mist-vibration coupling effects, and improves the environmental adaptability and the battery life of the marine UPS.

Description

Technical Field

[0001] This invention relates to the field of marine electrical automation technology, specifically to a PLC-based intelligent control system for marine UPS. Background Technology

[0002] Marine UPS (Uninterruptible Power Supply) devices are core equipment for ensuring the continuous and stable operation of critical loads on ships under conditions such as main power failure. They possess marine-specific characteristics such as double-conversion online topology, zero-interruption switching, wide voltage adaptability, high reliability, vibration resistance, and corrosion resistance. Traditional marine UPS systems often employ control methods based on fixed threshold switching or simple proportional-integral-derivative (PID) control, which have prominent problems: a lack of real-time perception and integration of environmental parameters; limited battery status monitoring methods, leading to continuous amplification of accumulated errors over long-term operation; edge-side data transmission only, lacking local preprocessing and intelligent analysis capabilities; and open-loop or semi-open-loop control strategies, failing to provide real-time feedback on environmental changes and battery degradation to control parameters.

[0003] In marine UPS applications, the actual degradation rate of batteries exhibits a significant nonlinear correlation with the product of salt spray concentration and vibration intensity, with a coefficient of determination far exceeding that with single environmental factors. The combined accelerated aging caused by salt spray and vibration further reveals that while the corrosion product layer formed by salt spray on the battery terminal surface provides some protection under static conditions, under continuous vibration stress from the ship, this corrosion layer continuously generates microcracks, exposing fresh metal surfaces and triggering a cycle of "corrosion-vibration-accelerated corrosion." Existing technologies have long treated temperature, salt spray, and vibration as independent factors, employing simple linear superposition corrections, failing to identify or incorporate the aforementioned coupling effects. This leads to severely inaccurate battery life predictions, delayed protection strategies, and a significantly increased risk of sudden failure.

[0004] Therefore, the existing technology lacks an intelligent control system that can specifically identify and quantitatively characterize the salt spray and vibration coupling acceleration effect for marine UPS applications, and deeply embed this coupling effect into the entire link of state estimation and control decision-making. Summary of the Invention

[0005] This invention overcomes the shortcomings of existing technologies and provides a PLC-based intelligent control system for marine UPS. Addressing the problems in existing marine UPS control technologies, such as the lack of identification and modeling of salt spray-vibration coupling acceleration effect, lack of environmental perception, insufficient state estimation accuracy, and weak control closed-loop feedback, this invention constructs a complete system integrating multi-source environmental perception, edge computing, coupled stress feature extraction, joint battery state estimation, and PLC adaptive control. The system acquires multi-dimensional information such as temperature, humidity, salt spray concentration, vibration acceleration, and battery electrical parameters in real time by deploying multi-source sensors in and around the UPS. In the edge computing unit, timestamp synchronization, anomaly removal, extended Kalman filtering for noise reduction, and data completion are performed to ensure the integrity and reliability of the input data. Salt spray-vibration coupling stress characteristics are extracted from the salt spray concentration and vibration acceleration data. The battery state estimation module uses an adaptive fusion filtering algorithm with coupled stress correction to jointly estimate the battery's state of charge and health status online. The process noise covariance matrix and the internal resistance aging term are linked through the same coupled stress sensitivity coefficient. The PLC main control unit integrates environmental characteristics, coupled stress characteristics, and state estimation results to dynamically execute nonlinear multi-level maintenance correction of the charging current and power supply mode switching decisions. The system combines data-driven learning with corrosion mechanism modeling through a "full-link connection of salt spray-vibration coupling effect + bidirectional feedback adaptive enhancement" mechanism. This enables coupled stress characteristics to simultaneously serve the improvement of state estimation accuracy and dynamic optimization of control strategies. Furthermore, by optimizing the feature extraction weights through execution feedback, a synergistic closed loop of "perception-coupled feature extraction-state estimation correction-control decision-execution feedback" is formed, significantly improving the adaptability, power supply reliability, and battery life of marine UPS in harsh marine environments.

[0006] The technical solution adopted by this invention is as follows: This solution provides a PLC-based intelligent control system for marine UPS, including an environmental perception layer, a battery perception layer, an edge computing unit, a battery state estimation module, a PLC main control unit, and an execution and feedback module.

[0007] The environmental sensing layer collects environmental data from the ship's engine room, including temperature, relative humidity, salt spray concentration, and vibration acceleration.

[0008] The battery sensing layer collects operating data of the battery packs in the UPS.

[0009] The edge computing unit receives environmental and operational data, performs timestamp synchronization, anomaly removal, noise reduction and smoothing, and packet loss compensation to obtain preprocessed operational data. It extracts environmental features from the environmental data, salt spray-vibration coupling stress features from the salt spray concentration and vibration acceleration data, and electrical features from the preprocessed operational data. The edge computing unit sends the preprocessed operational data and coupling stress features to the battery state estimation module, and sends the environmental, electrical, and coupling stress features to the PLC main control unit.

[0010] The battery state estimation module receives preprocessed operating data and coupled stress characteristics, and uses an adaptive fusion filtering algorithm with coupled stress correction to jointly estimate the battery's state of charge and health online, obtaining the state estimation result, which is then sent to the PLC main control unit in real time. The process noise covariance matrix and the internal resistance aging term in the battery equivalent circuit model are corrected by the same coupled stress sensitivity coefficient.

[0011] The PLC main control unit receives environmental characteristics, electrical characteristics, coupling stress characteristics output by the edge computing unit, and state estimation results output by the battery state estimation module. Based on the battery's effective state of charge (SOC) and a preset standard charging curve in the SOC estimation results, it determines the base charging current. Then, based on the battery health status in the SOC estimation results, the temperature compensation factor determined from temperature data in the environmental characteristics, and the coupling stress attenuation factor determined from the coupling stress characteristics, it performs multi-level correction on the base charging current to generate a charging current reference value. The PLC main control unit executes intelligent charging and discharging control strategies and dynamic power supply mode switching decisions, outputting control commands to the UPS rectifier, inverter, and charging circuit based on the charging current reference value. The coupling stress attenuation factor uses a nonlinear function, causing the correction amplitude to increase superlinearly when the coupling stress exceeds a preset threshold. The exponential parameter of the nonlinear function is determined through actual ship operation data.

[0012] The execution and feedback module receives control commands from the PLC main control unit, drives the power electronic devices of the UPS to perform corresponding charging and discharging actions and mode switching, and obtains actual execution parameters from the operation feedback of the power electronic devices; feeds the actual execution parameters back to the PLC main control unit, and at the same time obtains the actual charging and discharging depth and temperature change of the battery based on the operation data collected by the battery sensing layer, and feeds it back to the edge computing unit to optimize the extraction weight of coupling stress features.

[0013] Furthermore, the edge computing unit extracts salt spray-vibration coupled stress characteristics from salt spray concentration and vibration acceleration data. Specifically, this includes: performing sliding window cross-correlation analysis on the salt spray concentration sequence and vibration acceleration sequence to obtain the time-delay correlation coefficient between salt spray concentration changes and vibration events; performing rainflow counting processing on the vibration acceleration waveform to obtain the number of cycles at each stress amplitude level, and calculating the cumulative vibration fatigue damage based on the fatigue cumulative damage criterion; constructing a coupling acceleration factor based on the time-delay correlation coefficient and the cumulative vibration fatigue damage, which includes a material correlation index and a coupling coefficient; and fusing the coupling acceleration factor with salt spray concentration and vibration acceleration to generate coupled stress characteristics. The value ranges of the material correlation index and the coupling coefficient are determined through calibration using salt spray-vibration composite accelerated aging tests on marine valve-regulated sealed lead-acid batteries.

[0014] Furthermore, the battery state estimation module employs an adaptive fusion filtering algorithm with coupled stress correction to jointly estimate the battery's state of charge (SOC) and health state online. Specifically, this includes: establishing an equivalent circuit model for each battery cell; introducing a coupled stress correction term based on coupled stress characteristics into the internal resistance aging term of the equivalent circuit model; estimating the battery SOC based on an extended Kalman filter algorithm, where the process noise covariance matrix is ​​adaptively scaled according to coupled stress characteristics, and the correction coefficient of the process noise covariance matrix and the coupled stress sensitivity coefficient of the internal resistance aging term satisfy a preset proportional constraint relationship; simultaneously estimating the battery health state based on a dual Kalman filter architecture, introducing a coupled stress accelerated degradation term based on coupled stress characteristics into the state transition equation of the health state; and comparing the SOC of each cell in the battery pack and taking the minimum value as the effective SOC of the battery pack.

[0015] Furthermore, the PLC main control unit executes an intelligent charging and discharging control strategy, specifically including: performing multi-factor correction on the battery pack's effective state of charge, battery health status, temperature compensation factor, and coupling stress attenuation factor, and dynamically calculating the optimal charging current reference value; in the multi-factor correction, the temperature compensation factor and coupling stress attenuation factor act on the base charging current in a product form; the coupling stress attenuation factor takes a value of 1 when the coupling stress characteristic is lower than a preset first threshold, and attenuates in a power function form after the coupling stress characteristic exceeds the preset first threshold. The exponent parameter of the power function is determined by least squares identification through actual ship operation data, which is optimized for marine UPS batteries under real navigation conditions, and is different from the linear or logarithmic attenuation form commonly used in land-based fixed energy storage systems.

[0016] Furthermore, after the execution and feedback module feeds back the actual charge / discharge depth and temperature changes of the battery to the edge computing unit, the edge computing unit adjusts the extraction weights of the coupling stress features. Specifically, it calculates the battery temperature change rate based on the temperature change, and when the battery temperature change rate exceeds a preset threshold, it increases the weight coefficient of the vibration component in the coupling stress features. It also obtains the actual change rate of the battery internal resistance based on the operating data collected by the battery sensing layer, and when the deviation between the actual change rate of the battery internal resistance and the internal resistance change rate in the state estimation result exceeds a preset threshold, it uses the gradient descent method to adjust the material correlation index and coupling coefficient in the coupling acceleration factor, so that the internal resistance change rate gradually approaches the actual change rate of the battery internal resistance.

[0017] Compared with the prior art, the beneficial effects of the present invention are:

[0018] (1) The concentration of salt spray and vibration acceleration are collected in real time through the environmental perception layer. The sliding window cross-correlation analysis and rain flow counting are performed by the edge computing unit. The coupling acceleration factor is constructed and the salt spray-vibration coupling stress characteristics are generated. The nonlinear acceleration law of the coupling effect of salt spray and vibration on battery life is quantitatively characterized. This solves the problem of inaccurate life prediction caused by the existing technology that regards salt spray and vibration as independent factors and uses linear superposition correction.

[0019] (2) By using the battery state estimation module, the adaptive fusion filtering algorithm with coupling stress correction is adopted. The coupling stress correction term is introduced into the internal resistance aging term, and the process noise covariance matrix and the correction coefficient of the internal resistance aging term meet the preset proportional constraint relationship. This avoids the divergence of state estimation caused by independent parameter adjustment, significantly improves the estimation accuracy of battery state of charge and health under harsh conditions, and solves the problem of continuous amplification of state estimation cumulative error in the prior art.

[0020] (3) The PLC main control unit performs nonlinear multi-correction on the basic charging current based on the battery health status, temperature compensation factor and coupling stress attenuation factor. The coupling stress attenuation factor increases the correction amplitude in a superlinear manner after the coupling stress exceeds the threshold. The exponential parameter is dynamically identified through real ship data, so that the charging strategy matches the environmental stress level in real time. This solves the problem of open-loop control strategy and inability to provide real-time feedback on environmental changes and battery degradation in the existing technology.

[0021] (4) The actual execution parameters are fed back to the PLC main control unit through the execution and feedback module, and the actual charge and discharge depth and temperature change of the battery are fed back to the edge computing unit. The vibration weight coefficient and model parameters of the coupled stress characteristics are adjusted to form a synergistic closed loop of "sensing-feature extraction-state estimation-control decision-execution feedback". This solves the problem that the edge side of the existing technology has no local preprocessing and intelligent analysis capabilities, and significantly improves the environmental adaptability of the system and the battery life. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the overall structure of a PLC-based intelligent control system for marine UPS proposed in this invention.

[0023] Figure 2 This is a schematic diagram of the process for extracting salt spray-vibration coupled stress features by the edge computing unit in an embodiment of the present invention;

[0024] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. Detailed Implementation

[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0026] Example 1:

[0027] Please see Figure 1 This embodiment presents a PLC-based intelligent control system for marine UPS, applied to the online monitoring and intelligent control of marine UPS uninterruptible power supply devices. The system includes an environmental perception layer, a battery perception layer, an edge computing unit, a battery state estimation module, a PLC main control unit, and an execution and feedback module.

[0028] The environmental sensing layer collects environmental data from the ship's engine room and is deployed outside the UPS cabinet and its surrounding space. It is equipped with multi-source environmental sensors, including air temperature sensors, humidity sensors, atmospheric salt spray concentration sensors, cabinet surface condensation sensors, and a triaxial vibration acceleration sensor, to collect data on temperature, relative humidity, salt spray concentration, and vibration acceleration. The triaxial vibration acceleration sensor outputs acceleration time-domain waveforms in three orthogonal directions, with a sampling rate of 200Hz to capture the high-frequency components of ship vibrations.

[0029] The battery sensing layer collects operating data from the battery banks in the UPS. It is installed on the individual battery bank connection busbars and inside the battery cabinet, and is equipped with battery monitoring sensors, including voltage sensors, current sensors, internal resistance detection modules, and temperature sensors. The battery type is a marine valve-regulated sealed lead-acid battery. The operating data is aggregated to the edge computing unit via the Modbus communication protocol.

[0030] The edge computing unit uses an industrial-grade embedded computer installed inside the UPS cabinet. It receives environmental and operational data, performs preprocessing, including timestamp synchronization, anomaly removal, noise reduction and smoothing, and packet loss compensation, to obtain preprocessed operational data. It extracts environmental features from the environmental data, salt spray-vibration coupling stress features from the salt spray concentration and vibration acceleration data, and electrical features from the preprocessed operational data. The preprocessed operational data and coupling stress features are sent to the battery state estimation module, and the environmental, electrical, and coupling stress features are sent to the PLC main control unit.

[0031] The battery state estimation module receives preprocessed operating data and coupled stress characteristics, and uses an adaptive fusion filtering algorithm with coupled stress correction to jointly estimate the battery's state of charge and health state online, thus obtaining the state estimation result.

[0032] The PLC main control unit adopts a marine-grade programmable logic controller. It receives environmental characteristics, electrical characteristics, coupling stress characteristics output by the edge computing unit, and state estimation results output by the battery state estimation module. Based on the battery's effective state of charge in the state estimation results and the preset standard charging curve, it determines the basic charging current. Based on the battery health status in the state estimation results, the temperature compensation factor determined based on the temperature data in the environmental characteristics, and the coupling stress attenuation factor determined based on the coupling stress characteristics, it performs multi-level correction on the basic charging current to generate a charging current reference value. The PLC main control unit executes intelligent charging and discharging control strategies and dynamic power supply mode switching decisions, and outputs control commands to the rectifier, inverter, and charging circuit of the UPS based on the charging current reference value.

[0033] The execution and feedback module receives control commands from the PLC main control unit, drives the UPS power electronic devices to perform corresponding charging and discharging actions and mode switching, and obtains actual execution parameters from the operational feedback of the power electronic devices. Simultaneously, it acquires the actual charge / discharge depth and temperature changes of the battery based on operational data collected by the battery sensing layer. The execution and feedback module constitutes two key feedback loops: feeding back the actual execution parameters to the PLC main control unit to form a closed-loop correction of the inner control loop; and feeding back the actual charge / discharge depth and temperature changes of the battery to the edge computing unit to optimize the extraction weights of coupled stress features.

[0034] Example 2:

[0035] The data preprocessing process of the edge computing unit specifically includes the following steps:

[0036] Step A1: Timestamp Synchronization. The edge computing unit has a built-in real-time clock module that adds a unified timestamp to the data frames uploaded by the multi-source environmental sensors and battery monitoring sensors to obtain timestamp-synchronized data frames; using the time reference of the current and voltage data collected by the current and voltage sensors as the main time axis, the environmental data is time-interpolated and aligned.

[0037] Step A2: Outlier Detection and Removal. The Isolation Forest algorithm is used to detect outliers in the collected data frames after timestamp synchronization, resulting in data after removing outliers.

[0038] Step A3: Extended Kalman Filter Denoising. The data after outlier removal is denoised and smoothed using the extended Kalman filter algorithm to obtain the denoised data.

[0039] Step A4: Packet Loss Data Imputation. The system detects whether there is missing data due to packet loss in the denoised data. A regression prediction method based on historical trends is used to impute the missing data, resulting in the imputed data.

[0040] Step A5: Environmental and Electrical Feature Extraction. The completed data is divided into environmental data and battery operation data, which are used as preprocessed environmental data and preprocessed battery data, respectively. From the preprocessed environmental data, the average cabin temperature, temperature change rate, relative humidity, humidification time, daily average salt spray concentration, and vibration acceleration are extracted. From the preprocessed battery data, the average voltage, voltage range, current integral value, and temperature distribution characteristics of each cell are extracted.

[0041] Step A6: Salt spray-vibration coupled stress feature extraction. Please refer to [link / reference]. Figure 2 Specifically, it includes the following sub-steps:

[0042] Sub-step A6-1: Sliding window cross-correlation analysis. This involves analyzing the salt spray concentration sequence. and vibration acceleration sequence Perform sliding window cross-correlation analysis. Let the window length be... Calculate the salt spray concentration change within each window. The time-delay correlation coefficient between the vibration event and the vibration event. A vibration event is defined as vibration acceleration exceeding a preset threshold. Calculate the time-delay correlation coefficient. :

[0043] ;

[0044] in, This represents the number of sampling points within the sliding window; The first in the salt spray concentration sequence Salt spray concentration values ​​at each sampling point; This represents the mean of the salt spray concentration sequence within the sliding window. ; The first in the vibration acceleration sequence Vibration acceleration values ​​at each sampling point; This represents the mean of the vibration acceleration sequence within the sliding window. .

[0045] Sub-step A6-2: Calculation of cumulative vibration fatigue damage. Perform rainflow counting processing on the vibration acceleration waveform to obtain the number of cycles at each stress amplitude level. The cumulative damage amount of vibration fatigue was calculated using the Palmgren-Miner linear cumulative damage criterion. :

[0046] ;

[0047] in, The fatigue life cycle number of the material under this stress amplitude is determined by the SN curve of the battery terminal material.

[0048] Sub-step A6-3: Construction of the coupling acceleration factor. Based on the time-delay correlation coefficient. and cumulative damage from vibration fatigue Constructing a coupling acceleration factor :

[0049] ;

[0050] in, The average salt spray concentration within the sliding window. For reference salt spray concentration, For reference damage amount, and For material-related indices, The coupling coefficient is used. The range of values ​​for the material correlation index and the coupling coefficient is determined through calibration using salt spray-vibration accelerated aging tests on marine valve-regulated sealed lead-acid batteries.

[0051] The specific calibration method is as follows: Orthogonal experiments were conducted on 36 marine valve-regulated sealed lead-acid battery samples in a salt spray-vibration composite environmental test chamber. Salt spray concentrations were set at four levels: 0.2, 0.6, 1.0, and 1.4 mg / m³; vibration accelerations were set at four levels: 0.5, 1.5, 2.5, and 3.5 m / s²; and the test period was 60 days. The terminal corrosion depth and internal resistance increment of each sample were measured every 10 days. The experimental data were fitted using a coupled acceleration factor model, and the parameters were optimized using response surface methodology. For the lead-calcium alloy terminal material of the marine valve-regulated sealed lead-acid battery, the material correlation index... The optimal value range is 0.62 to 0.78. The optimal value range for the coupling coefficient is 0.58 to 0.72. The optimal value range is 0.28 to 0.45. This range differs significantly from the parameter values ​​commonly used in terrestrial stationary energy storage systems (in terrestrial systems). Typically, it is taken as 0.4 to 0.5. The concentration is usually taken as 0.3 to 0.4. This is because the chloride ion concentration and vibration spectrum characteristics of salt spray in marine environments are fundamentally different from those in terrestrial environments. Marine vibrations are mainly low-frequency, large-amplitude wave impacts and main engine vibrations, while terrestrial energy storage vibrations are mainly high-frequency, small-amplitude traffic or equipment vibrations. The contribution mechanisms of the two to corrosion layer rupture are different.

[0052] Sub-step A6-4: Coupling stress feature fusion. This involves fusing the coupling acceleration factor. Salt spray concentration at the current moment Vibration acceleration Fusion is performed to generate salt spray-vibration coupled stress characteristics. :

[0053] ;

[0054] in, For reference vibration acceleration, and These are the salt spray weighting coefficient and the vibration weighting coefficient, both initially set to 0.5. The weighting coefficients are dynamically adjusted based on the actual depth of charge / discharge and temperature changes of the battery returned by the execution and feedback module.

[0055] Example 3:

[0056] The specific workflow of the battery state estimation module is as follows:

[0057] Step B1: Equivalent Circuit Model Establishment and Coupling Stress Correction. For valve-regulated sealed lead-acid batteries, a first-order equivalent circuit model is established, consisting of an open-circuit voltage source. Ohmic internal resistance and polarization resistance With polarization capacitor The RC circuit, composed of parallel connections, is connected in series. This determines the battery's state of charge. polarization voltage across the polarization capacitor as a state vector , will charge and discharge current The battery terminal voltage is used as the input variable and the output variable. Establish the discrete state-space equations:

[0058] ;

[0059] ;

[0060] in, Here is the state transition matrix. , The polarization time constant is The sampling time interval; For the input matrix, , For Coulomb efficiency, This refers to the battery's rated capacity. This is process noise; For measuring noise; The nonlinear observation function represents the open-circuit voltage source. With state of charge The nonlinear mapping relationship between them.

[0061] In the internal resistance of Ohm Introducing a coupling stress feature-based aging evolution model Constructed coupling stress correction term:

[0062] ;

[0063] in, The initial internal resistance, Based on the aging rate coefficient, This is the coupling stress sensitivity coefficient. Calibrated through a composite accelerated aging test, for marine valve-regulated sealed lead-acid batteries, The typical value range is 0.08 to 0.15.

[0064] Step B2: Adaptive State of Charge Estimation with Coupled Stress Correction. The state of charge of each cell is estimated based on the extended Kalman filter algorithm. Process noise covariance matrix. Based on the characteristics of coupled stress Perform adaptive scaling:

[0065] ;

[0066] in, The basic process noise covariance matrix, This is the noise scaling factor. After calibration, The optimal value range is 0.8 to 1.8.

[0067] Scaling factor of process noise covariance matrix Coupling stress sensitivity coefficient with internal resistance aging term The relationships between them satisfy a preset proportional constraint. Specifically, under different operating conditions, and The ratio remains stable within the range of 10.5 to 12.0. This is because the coupling stress affects the uncertainty of the state estimation (manifested as...). The scaling effect and its impact on the physical aging rate (manifested as an increase in internal resistance) are of the same origin, both stemming from the intensified fluctuations in the internal electrochemical state of the battery caused by the salt spray-vibration synergistic effect. Therefore, in the algorithm implementation, Constraints Linear functions:

[0068] ;

[0069] in, This is a proportionality coefficient, ranging from 10.5 to 12.0. The introduction of this correlation constraint makes the two correction stages of the state estimation module form an organic whole, avoiding the state estimation divergence problem that might result from independent parameter adjustments.

[0070] Step B3: Joint estimation of the health state with coupling stress correction. A dual Kalman filter architecture is used to simultaneously estimate the battery health state. The state transition equation for the health state incorporates features based on coupling stress. Constructed coupled stress-accelerated degradation term:

[0071] ;

[0072] ;

[0073] in, for The parameter vector at time step, including the ohmic internal resistance. and battery rated capacity ; The basic degradation is the natural degradation value of the parameter in each sampling period under uncoupled stress. This refers to the additional degradation caused by coupling stress; For parameter process noise; The coupling stress degradation coefficients for each parameter; It is a non-linear exponent, after calibration. The optimal range of values ​​is to This reflects the superlinear characteristics of accelerated degradation under high stress; for The characteristic value of the coupled stress at time t; This represents the sampling time interval.

[0074] Battery health status is expressed as available capacity in the parameter vector. With battery rated capacity The ratio represents:

[0075] ;

[0076] in, for Time parameter vector The available capacity value in the table is the current available capacity.

[0077] Step B4: Determine the effective state of charge (SOC) of the battery pack. This is done by comparing the SOC of each individual cell in the battery pack. The minimum value is taken as the effective state of charge of the battery pack. :

[0078] ;

[0079] in, This represents the total number of individual cells connected in series in the battery pack. For the first Estimated state of charge of individual cells.

[0080] Step B5: Output the state estimation results. Output the state of charge (SOC) of each individual cell and the effective SOC of the battery pack. Internal resistance of each individual cell, battery health status The pressure difference between individual units is used as the state estimation result and sent to the PLC main control unit in real time.

[0081] Example 4:

[0082] The workflow of the PLC main control unit is as follows:

[0083] Step C1: Input Data Integration. Receive the average cabin temperature output from the edge computing unit within each control cycle. Coupling stress characteristics The battery pack's effective state of charge is output by the battery state estimation module. and battery pack health status ,Will , , , Together they constitute the input data for decision-making.

[0084] Step C2: Multi-level calibration calculation of charging current. Based on the decision input data, perform multi-level calibration on the base charging current to dynamically calculate the optimal charging current reference value. :

[0085] ;

[0086] in, To be based on the standard three-stage charging curve and the effective state of charge of the battery pack Determined base charging current; This is a temperature compensation factor, based on the average cabin temperature. It is determined that a Gaussian decay form will be used, with a value of 1 at the optimal temperature. This is a health status degradation factor, based on the battery pack's health status. It is determined that an S-shaped function is used, and the correction coefficient drops rapidly when the battery is in a healthy state and falls to near the threshold, thereby reducing the charging current to protect the battery during aging. This is the coupling stress attenuation factor, based on the characteristics of salt spray-vibration coupled stress. It is confirmed that the charging current is actively limited under high coupling stress conditions. Using a nonlinear piecewise function form:

[0087] ;

[0088] in, Low stress threshold, For high stress threshold, The attenuation coefficient is... It is a non-linear exponent. This is the lower limit of the minimum correction coefficient. Nonlinear exponent. Dynamically determined through least-squares identification using actual ship operational data. Specifically, the identification method involves collecting 30 days of continuous operational data during the ship's voyage, including coupled stress characteristics. The time series of the battery's actual capacity decay rate and the time series of the battery's actual capacity decay rate were obtained. The capacity decay rate was fitted with coupling stress characteristics using nonlinear regression, and the Levenberg-Marquardt algorithm was used to optimize the solution with the objective of minimizing the sum of squared residuals. The optimal value was determined by analyzing real-ship data from five vessels on different routes. The optimal value range is 1.5 to 2.5, and this value is significantly correlated with the climate characteristics of the main navigation areas of ships. The optimal value for tropical sea routes is... Values ​​are concentrated between 2.0 and 2.5, indicating the optimal route for temperate sea areas. The values ​​are concentrated between 1.5 and 2.0. Marine UPS control strategies must dynamically identify the nonlinear exponent based on actual navigation conditions to achieve optimal battery life protection. In contrast, land-based stationary energy storage systems commonly use linear or logarithmic decay forms (equivalent to...). (Or logarithmic functions) cannot match this superlinear acceleration feature at all.

[0089] At the same time, based on the battery pack health status Salt spray-vibration coupled stress characteristics Determine the discharge protection threshold, including the discharge cutoff voltage and the maximum discharge current.

[0090] Step C3: Dynamic adjustment of discharge protection threshold. The discharge cutoff voltage and maximum discharge current are dynamically adjusted based on the battery pack's health status and coupling stress characteristics.

[0091] Step C4: Dynamic power supply mode switching decision. When the coupling stress characteristics exceed the preset threshold, the system will actively limit the upper limit of discharge power even if the battery's state of charge is still acceptable.

[0092] Step C5: Early Warning and Protection. When the coupling stress characteristics exceed the safety threshold, an environmental protection early warning is issued and the charging and discharging power is actively reduced; when the rate of increase of the coupling stress characteristics exceeds the preset rate threshold, the system automatically enters the protective operation mode.

[0093] Step C6: Control command generation. The charging current reference value is converted into a rectifier firing angle command and a charging circuit pulse width modulation signal, and sent to the execution and feedback module.

[0094] Example 5:

[0095] The weights for extracting coupling stress features are optimized in edge computing units. The specific steps are as follows:

[0096] Step D1: Calculate the battery temperature change rate based on the battery temperature changes reported by the execution and feedback module. When the battery temperature change rate exceeds a preset threshold, it is determined that the battery is experiencing increased internal heating under the current operating conditions. This is related to increased contact resistance caused by internal structural fretting friction due to vibration or salt spray corrosion. At this point, the weighting coefficient of the vibration component in the coupled stress characteristics is automatically increased. Updated vibration weighting coefficients and the updated salt spray weighting coefficient They are respectively:

[0097] ;

[0098] ;

[0099] in, Adjust the step size for the preset weights; This is the preset upper limit value for the vibration weighting coefficient.

[0100] Step D2: Obtain the actual rate of change of battery internal resistance based on the operational data collected by the battery sensing layer. When the deviation between the actual rate of change of battery internal resistance and the rate of change of internal resistance in the state estimation result output by the battery state estimation module exceeds a preset threshold, adjust the material-related index in the coupling acceleration factor using the gradient descent method. , and coupling coefficient This allows the rate of change of internal resistance in the state estimation results to gradually approximate the actual rate of change of battery internal resistance. The updated coupling coefficients... for:

[0101] ;

[0102] in, The preset learning rate; The mean square error loss function is the difference between the rate of change of internal resistance in the state estimation results and the actual rate of change of battery internal resistance. For loss function For coupling coefficients Partial derivatives. Material-related indices. and The adjustment adopts and A similar gradient descent method.

[0103] The introduction of this reverse feedback path allows the feature extraction stage to move beyond fixed feedforward calculations and instead adaptively optimize based on the battery's actual operating response. As the system's operating time accumulates, the coupled stress feature model increasingly approximates the real physical degradation mechanism, and the accuracy of state estimation and control strategies continues to improve.

[0104] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A PLC-based intelligent control system for marine UPS, characterized in that, It includes an environmental perception layer, a battery perception layer, an edge computing unit, a battery state estimation module, a PLC main control unit, and an execution and feedback module. The environmental perception layer collects environmental data from the ship's engine room, including temperature, salt spray concentration, and vibration acceleration. The battery perception layer collects operating data from the battery pack in the UPS. The edge computing unit receives environmental data and operating data, preprocesses them to obtain preprocessed operating data, extracts environmental features from the environmental data, extracts salt spray-vibration coupling stress features from the salt spray concentration and vibration acceleration, and extracts electrical features from the preprocessed operating data. The battery state estimation module performs joint online estimation of the battery's state of charge and state of health to obtain the state estimation result; The PLC main control unit determines the base charging current based on the battery's effective state of charge in the state estimation results and the preset standard charging curve, performs multiple corrections on the base charging current, generates a charging current reference value, and outputs control commands. The execution and feedback module receives control commands, drives the UPS power electronic devices to perform charging and discharging actions and mode switching, obtains actual execution parameters from the operation feedback of the power electronic devices, and feeds them back to the PLC main control unit; it obtains the actual charging and discharging depth and temperature changes of the battery based on the operation data, and feeds them back to the edge computing unit to optimize the extraction weight of coupling stress features.

2. The intelligent control system for marine UPS based on PLC control according to claim 1, characterized in that: The edge computing unit extracts the salt spray-vibration coupled stress characteristics, specifically including: performing sliding window cross-correlation analysis on salt spray concentration and vibration acceleration to obtain the time-delay correlation coefficient; performing rainflow counting processing on vibration acceleration to obtain the number of cycles under each stress amplitude, and calculating the vibration fatigue cumulative damage based on the fatigue cumulative damage criterion; constructing a coupling acceleration factor based on the time-delay correlation coefficient and the vibration fatigue cumulative damage; and fusing the coupling acceleration factor with salt spray concentration and vibration acceleration to generate coupled stress characteristics.

3. The intelligent control system for marine UPS based on PLC control according to claim 2, characterized in that: The battery state estimation module employs an adaptive fusion filtering algorithm with coupled stress correction to jointly estimate the battery's state of charge (SOC) and health state online. Specifically, this includes: establishing an equivalent circuit model for each battery cell; introducing a coupled stress correction term based on coupled stress characteristics into the internal resistance aging term of the equivalent circuit model; estimating the SOC based on an extended Kalman filter algorithm; wherein the process noise covariance matrix in the coupled stress characteristic correction adaptive fusion filtering algorithm and the internal resistance aging term in the battery equivalent circuit model are adaptively scaled according to coupled stress characteristics; simultaneously estimating the battery health state based on a dual Kalman filter architecture; introducing a coupled stress accelerated degradation term based on coupled stress characteristics into the health state's state transition equation; and comparing the SOC of each cell in the battery pack, taking the minimum value as the battery's effective SOC.

4. The intelligent control system for marine UPS based on PLC control according to claim 3, characterized in that: The correction coefficient of the process noise covariance matrix and the coupling stress sensitivity coefficient of the internal resistance aging term satisfy the preset proportional constraint relationship.

5. The intelligent control system for marine UPS based on PLC control according to claim 3, characterized in that: The PLC main control unit performs multiple corrections on the base charging current, specifically including: determining the temperature compensation factor based on the temperature in the environmental characteristics; determining the coupling stress attenuation factor based on the coupling stress characteristics; performing multiple corrections on the battery's effective state of charge, battery health state, temperature compensation factor, and coupling stress attenuation factor, and dynamically calculating the charging current reference value; and applying the temperature compensation factor and coupling stress attenuation factor to the base charging current in a product form.

6. The intelligent control system for marine UPS based on PLC control according to claim 5, characterized in that: The coupling stress attenuation factor is 1 when the coupling stress characteristic is below the preset first threshold, and attenuates in the form of a power function after the coupling stress characteristic exceeds the preset first threshold.

7. The intelligent control system for marine UPS based on PLC control according to claim 5, characterized in that: After the execution and feedback module feeds back the actual charge and discharge depth and temperature change of the battery to the edge computing unit, the edge computing unit calculates the battery temperature change rate based on the temperature change. When the battery temperature change rate is detected to exceed the preset threshold, the weighting coefficient of the vibration component in the coupled stress feature is increased.

8. The intelligent control system for marine UPS based on PLC control according to claim 7, characterized in that: After the execution and feedback module feeds back the actual charge and discharge depth and temperature changes of the battery to the edge computing unit, the edge computing unit obtains the actual change rate of the battery internal resistance based on the running data. When the deviation between the actual change rate of the battery internal resistance and the internal resistance change rate in the state estimation result exceeds the preset threshold, the gradient descent method is used to adjust the material-related index and coupling coefficient in the coupling acceleration factor so that the internal resistance change rate gradually approaches the actual change rate of the battery internal resistance.

Images