Supercritical organic rankine cycle efficiency optimization method and system

Abstract

This invention discloses a method and system for optimizing the efficiency of a supercritical organic Rankine cycle. It analyzes monitoring data to pinpoint stable operating segments and trace the starting point of performance degradation, extracting operational characteristics. It establishes working fluid state correlations and analyzes phase transition characteristics to generate a benchmark database. Through state matching, it identifies deviation types, analyzes the propagation path and attenuation law of deviations within the system, and quantifies the degree of performance deviation. It establishes response relationships between devices and quantifies dynamic coupling characteristics, capturing fluctuation response characteristics. Sensitive time periods are identified through envelope analysis and response intensity. Energy transfer analysis is performed on the flow path to calculate parameters at unmonitored locations and obtain performance distribution characteristics. The rate of change of performance deviation is extracted and combined with device coordination characteristics to predict sudden changes, determine response time difference and lead time, and generate graded adjustment commands based on performance distribution characteristics. This enables the prediction and optimized control of the efficiency of a supercritical organic Rankine cycle system.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of energy recovery and conversion, in particular to a supercritical organic Rankine cycle efficiency optimization method and system. BACKGROUND

[0002] As an advanced low-grade heat recovery technology, the supercritical organic Rankine cycle system has shown good application prospects in industrial waste heat utilization, geothermal power generation, solar thermal power generation and other fields. The system utilizes the special thermophysical properties of the working fluid in the supercritical state to realize the conversion of heat energy to mechanical energy, and has higher energy utilization rate and wider heat source temperature adaptation range than traditional cycle systems.

[0003] However, in the actual operation process of the supercritical organic Rankine cycle system, the system performance is easily affected by factors such as heat source fluctuations, environmental condition changes, equipment aging, and deviates from the design condition, resulting in continuous decline in operating efficiency. The existing optimization control method mainly focuses on the adjustment of a single device or a single parameter, lacks analysis of the evolution law of the overall running state of the system, and is difficult to identify and intervene at the early stage of performance degradation. At the same time, due to the complex dynamic coupling relationship between the internal devices of the system, the traditional adjustment strategy often responds slowly, and only takes measures when the efficiency is significantly reduced, missing the best intervention opportunity. SUMMARY

[0004] The present application discloses a supercritical organic Rankine cycle efficiency optimization method and system, which locates the performance degradation starting point by time series analysis of operation data, establishes a benchmark database to quantify the current state deviation, analyzes the dynamic response relationship between devices to identify key influence periods, calculates the running state of unmonitored positions to obtain performance distribution characteristics, predicts performance mutation risks and issues adjustment instructions in advance, and realizes intelligent optimization control of the supercritical organic Rankine cycle system efficiency.

[0005] The first aspect of the present application proposes a supercritical organic Rankine cycle efficiency optimization method, comprising the following steps:

[0006] Collecting multi-device monitoring data in the operation of the supercritical organic Rankine cycle system, performing working condition deviation tracing processing on the multi-device monitoring data to generate cycle operation characteristics;

[0007] Establishing a working fluid state correlation table according to the cycle operation characteristics, identifying phase state response from supercritical region pressure-temperature pairing of the working fluid state correlation table, extracting phase change boundary parameters from the phase state response to generate an efficiency benchmark database, and performing current working condition deviation detection based on the efficiency benchmark database to generate an efficiency deviation measure;

[0008] Based on the cyclic operation characteristics, load matching analysis between equipment is performed to establish a cross-equipment response chain. The time-series coupling degree between the expander and the evaporator is determined through the cross-equipment response chain. The time-series coupling degree is used to capture load fluctuation response characteristics, and the efficiency-sensitive interval is extracted from the load fluctuation response characteristics.

[0009] Within the efficiency-sensitive range, the efficiency deviation measurement is amplified to generate equipment collaborative features. The equipment collaborative features are used to perform working fluid circulation path tracking to generate a flow path index. The flow path index is analyzed by specific enthalpy-specific entropy transfer to generate unmeasured parameters. The heat transfer efficiency change features are captured from the unmeasured parameters to obtain efficiency cascade data.

[0010] The efficiency deviation metric and the equipment coordination characteristics are used to predict efficiency jumps and determine the adjustment response time difference. Based on the adjustment response time difference and the efficiency tier data, a graded adjustment command is generated.

[0011] A second aspect of this invention proposes a supercritical organic Rankine cycle efficiency optimization system, comprising:

[0012] The data acquisition module is used to collect multi-device monitoring data during the operation of the supercritical organic Rankine cycle system, and to perform operating condition offset tracing processing on the multi-device monitoring data to generate cycle operation characteristics;

[0013] The efficiency evaluation module is used to establish a working fluid state association table based on the cyclic operation characteristics, perform supercritical pressure-temperature pairing identification of the working fluid state association table to identify the phase response, extract phase transition limit parameters from the phase response to generate an efficiency benchmark database, and perform current operating condition deviation detection based on the efficiency benchmark database to generate an efficiency deviation metric.

[0014] The collaborative analysis module is used to perform load matching analysis between equipment to establish a cross-equipment response chain for the cyclic operation characteristics, determine the time coupling degree between the expander and the evaporator through the cross-equipment response chain, capture the load fluctuation response characteristics using the time coupling degree, and extract the efficiency sensitive interval from the load fluctuation response characteristics.

[0015] The parameter generation module is used to amplify the efficiency deviation measurement within the efficiency sensitive range to generate equipment collaborative features, perform working fluid circulation path tracing on the equipment collaborative features to generate a flow path index, perform specific enthalpy-specific entropy transfer analysis on the flow path index to generate unmeasured parameters, and capture heat transfer efficiency change features from the unmeasured parameters to obtain efficiency cascade data.

[0016] The instruction generation module is used to predict efficiency jumps based on the efficiency deviation metric and the equipment coordination characteristics to determine the adjustment response time difference, and generate graded adjustment instructions based on the adjustment response time difference and the efficiency tier data.

[0017] The beneficial effects of this invention are reflected in the following points: First, traditional optimization methods struggle to determine when system performance begins to deteriorate and where the root cause of this deterioration lies. This invention identifies the most stable period of system operation as a reference benchmark through fluctuation analysis of historical operating data, and traces back from this benchmark to find the starting point of the first performance degradation. Based on this, a working fluid state correlation is established and phase transition characteristics are analyzed. Phase transition boundary parameters are extracted to form a benchmark database. Matching the current operating state with the benchmark can identify the type and direction of deviation. Furthermore, by tracing the diffusion path of deviation among various devices within the system, the attenuation law during the deviation propagation process is quantified. The deviation propagation distance is used as a metric for performance deviation, accurately locating the starting moment of performance degradation and revealing the propagation mechanism of deviation. Second, combining the temporal coupling analysis between devices with the extraction of load fluctuation response characteristics can reveal the dynamic response law of the system to disturbances. Specifically, the moment of load change of key equipment is extracted as a trigger marker, the delay time of response of other equipment is monitored, and stability statistics are performed, thereby quantifying the degree of temporal coupling between devices. After understanding the temporal coupling characteristics, envelope decomposition and threshold crossing detection are performed on the load fluctuation response signal to identify sensitive fluctuation segments where the response amplitude exceeds the threshold. The response intensity is then determined by combining amplitude, duration, and energy accumulation, ultimately identifying the critical time period when the system is most sensitive to disturbances and clarifying the propagation sequence of disturbances between equipment. Finally, traditional methods rely on data from a limited number of monitoring points and often only adjust after a significant performance degradation. This invention establishes a node distribution along the working fluid flow path, applies energy conservation constraints to calculate the energy state of each node, and uses pressure information to back-calculate the thermodynamic parameters of the working fluid, thus obtaining a complete state distribution including unmonitored locations. The spatial distribution of heat transfer performance is extracted from the calculated parameters and classified into high and low levels, providing a basis for graded regulation. Simultaneously, a sudden change prediction mechanism is introduced, extracting the temporal change rate of performance deviation metrics, identifying rate inflection points based on equipment collaborative characteristics, assessing the risk of sudden changes, generating early warning signs, and determining the advance issuance time of regulation commands based on the warning time. Regulation commands of corresponding strength are generated based on performance distribution characteristics, enabling regulation measures to take effect before performance sudden changes occur. Attached Figure Description

[0018] The accompanying drawings illustrate specific examples of the technical solutions described in this invention and, together with the detailed embodiments, form part of the specification, serving to explain the technical solutions, principles, and effects of this invention.

[0019] Unless otherwise specified or otherwise, the same reference numerals in different figures represent the same or similar technical features, and different reference numerals may be used to represent the same or similar technical features.

[0020] Figure 1This is a schematic diagram of the process for optimizing the efficiency of the supercritical organic Rankine cycle of the present invention.

[0021] Figure 2 This is a structural block diagram of the supercritical organic Rankine cycle efficiency optimization method and system of the present invention. Detailed Implementation

[0022] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0023] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0024] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0025] The technical solutions of the embodiments of this application will be described below.

[0026] like Figure 1 As shown, this embodiment of the invention provides a method and system for optimizing the efficiency of a supercritical organic Rankine cycle, including the following steps S110-S150:

[0027] Step S110: Collect multi-device monitoring data during the operation of the supercritical organic Rankine cycle system, and perform operating condition offset tracing processing on the multi-device monitoring data to generate cycle operation characteristics.

[0028] Specifically, monitoring data from multiple devices in a supercritical organic Rankine cycle system were collected. Sensors were deployed on key devices of the supercritical organic Rankine cycle system, including the evaporator, expander, condenser, and working fluid pump. Inlet and outlet temperature sensors were deployed on the evaporator to monitor temperature changes of the working fluid entering and exiting the evaporator; inlet and outlet pressure sensors were deployed to monitor pressure changes of the working fluid entering and exiting the evaporator; and a heat source flow meter was deployed to monitor the flow rate of the heat source fluid. Inlet and outlet temperature and pressure sensors were deployed on the expander to monitor the inlet and outlet operating conditions; a speed sensor was deployed to monitor the expander speed; and a power sensor was deployed to monitor the expander output power. Inlet and outlet temperature and pressure sensors were deployed on the condenser to monitor the inlet and outlet operating conditions; and a cooling water flow meter was deployed to monitor the cooling water flow rate. Inlet and outlet pressure sensors were deployed on the working fluid pump to monitor the pump's pressurization effect; a flow meter was deployed to monitor the working fluid flow rate; and a power sensor was deployed to monitor the pump's input power. The sensor sampling frequency was set to 1Hz, and data from each sensor was recorded in real time. The monitoring data, including temperature, pressure, flow rate, and power, from each device are aggregated to form multi-device monitoring data. This data is then synchronized and aligned according to timestamps to ensure temporal consistency across different devices. The multi-device monitoring data includes all operating parameters for the evaporator, expander, condenser, and working fluid pump.

[0029] In some embodiments, the step of performing operating condition offset tracing processing on the multi-device monitoring data to generate cyclic operation features includes: performing efficiency fluctuation scanning on the multi-device monitoring data to obtain a fluctuation amplitude sequence; performing benchmark value positioning on the fluctuation amplitude sequence to generate a steady-state benchmark segment; using the steady-state benchmark segment as the tracing starting point to search backward and lock the first efficiency drop position; and extracting equipment status parameters based on the first efficiency drop position to construct cyclic operation features.

[0030] Efficiency fluctuations were scanned using multi-device monitoring data to obtain a fluctuation amplitude sequence. Based on the power data from the multi-device monitoring data, the net output power W_net = W_exp - W_pump was calculated, where W_exp is the expander output power and W_pump is the working fluid pump input power. According to the evaporator temperature and pressure from the multi-device monitoring data, the thermodynamic properties of the working fluid were consulted to obtain the evaporator inlet enthalpy h_in and outlet enthalpy h_out. Based on the flow rate and enthalpy from the multi-device monitoring data, the input heat Q_in = m × (h_out - h_in) was calculated, where m is the working fluid mass flow rate, h_out is the evaporator outlet enthalpy, and h_in is the evaporator inlet enthalpy. The instantaneous efficiency η = W_net / Q_in of the supercritical organic Rankine cycle system was calculated. A sliding time window was used to scan the efficiency calculated from the multi-device monitoring data, with a time window width of 60 seconds and a window sliding step size of 10 seconds. Within each time window, the average and standard deviation of the efficiency were calculated. The fluctuation amplitude A is defined as the ratio of the standard deviation to the mean: A = σ / η_avg × 100%, where σ is the standard deviation of efficiency within the time window, η_avg is the average efficiency within the time window, and the fluctuation amplitude reflects the degree of efficiency fluctuation within that time window. The fluctuation amplitudes of each time window are arranged in chronological order to form a fluctuation amplitude sequence. Each element of the fluctuation amplitude sequence corresponds to the degree of efficiency fluctuation within a time window; a smaller fluctuation amplitude indicates more stable operation during that period, while a larger fluctuation amplitude indicates more drastic efficiency fluctuations during that period.

[0031] A steady-state benchmark segment is generated by locating benchmark values ​​for the fluctuation amplitude sequence. The numerical distribution characteristics of the fluctuation amplitude sequence are analyzed to identify time periods with smaller fluctuation amplitudes. A steady-state determination threshold is set; when the fluctuation amplitude in the sequence is less than this threshold, the corresponding time period is considered a steady-state operating segment. The steady-state determination threshold is set to 3%, meaning that a fluctuation amplitude less than 3% is considered steady-state. The fluctuation amplitude sequence is traversed, and all time windows that meet the steady-state conditions are marked as steady-state time windows. Consecutive steady-state time windows in the fluctuation amplitude sequence are merged into steady-state intervals. The duration of each steady-state interval is calculated, and the steady-state interval with the longest duration is selected as the steady-state benchmark segment. The steady-state benchmark segment corresponds to the period when the supercritical organic Rankine cycle system operates most stably and with the smallest operational deviation; during this period, the efficiency fluctuation amplitude remains at a low level. The start and end times of the steady-state benchmark segment in the fluctuation amplitude sequence are recorded. Efficiency data for the corresponding time period of the steady-state benchmark segment are extracted, and the average efficiency of this segment is calculated as a benchmark value to quantify the degree to which the system deviates from steady-state operation.

[0032] The steady-state baseline is used as the starting point for tracing back to pinpoint the location of the first efficiency drop. Starting from the end of the steady-state baseline, the efficiency change is searched forward. Efficiency data for the time period following the steady-state baseline is extracted to construct an efficiency time-series curve. The deviation of efficiency at each moment from the average efficiency of the steady-state baseline is calculated as Δη(t) = η_baseline - η(t), where η_baseline is the average efficiency of the steady-state baseline, and η(t) is the instantaneous efficiency at time t. An efficiency drop threshold is set; when the efficiency deviation exceeds this threshold, it is considered an efficiency drop. The efficiency drop threshold is set to 5% of the average efficiency of the steady-state baseline, meaning that a drop is considered to occur when the efficiency decreases by more than 5% of the baseline value. Starting from the end of the steady-state baseline, the efficiency deviation is checked moment by moment. When the efficiency deviation at a certain moment first exceeds the drop threshold, that moment is marked as the location of the first efficiency drop. The location of the first efficiency drop corresponds to the critical moment when the operating condition of the supercritical organic Rankine cycle system begins to shift, marking the transition of the system from steady-state operation to a performance degradation state. Record the timestamp of the first efficiency drop, the corresponding instantaneous efficiency, and the efficiency deviation.

[0033] Based on the initial efficiency drop point, equipment state parameters are extracted to construct a cyclic operation feature. At the initial efficiency drop point, state parameters of each device are extracted to characterize the operating condition deviation. State parameters for the evaporator at the initial efficiency drop point are extracted: evaporator inlet temperature, evaporator outlet temperature, evaporator inlet pressure, and evaporator outlet pressure, and the evaporation temperature difference is calculated. State parameters for the expander at the initial efficiency drop point are extracted: expander inlet temperature, expander outlet temperature, expander speed, and expander output power. State parameters for the condenser at the initial efficiency drop point are extracted: condenser inlet temperature, condenser outlet temperature, and condenser inlet pressure. State parameters for the working fluid pump at the initial efficiency drop point are extracted: working fluid flow rate, pump outlet pressure, and pump input power. The equipment state parameters extracted at the initial efficiency drop point are integrated into a cyclic operation feature, including 5 parameters for the evaporator, 4 parameters for the expander, 3 parameters for the condenser, and 3 parameters for the working fluid pump, totaling 15 feature dimensions. The cyclic operation characteristics fully describe the operating state of the supercritical organic Rankine cycle system at the moment of the first efficiency drop, providing a data foundation for subsequent operating condition offset analysis.

[0034] Step S120: Establish a working fluid state association table based on the cyclic operation characteristics, identify the phase response by supercritical pressure-temperature pairing on the working fluid state association table, extract phase transition boundary parameters from the phase response to generate an efficiency benchmark database, and generate an efficiency deviation metric based on the efficiency benchmark database to detect the deviation of the current operating condition.

[0035] Specifically, a working fluid state correlation table is established based on the characteristics of the cyclic operation. The evaporator inlet temperature, evaporator outlet temperature, evaporator inlet pressure, and evaporator outlet pressure are marked as state points 1 and 2, respectively. The expander inlet temperature and expander inlet pressure are designated as state point 3, and the expander outlet temperature and expander outlet pressure as state point 4. The expander inlet corresponds to the evaporator outlet, and state point 3 is consistent with state point 2. The condenser inlet temperature and condenser inlet pressure are designated as state point 5, and the condenser outlet temperature and condenser outlet pressure as state point 6. The condenser inlet corresponds to the expander outlet, and state point 5 is consistent with state point 4. The condenser outlet, as state point 6, is also the working fluid pump inlet; the working fluid pump outlet is designated as state point 7, and the working fluid pump outlet corresponds to the evaporator inlet. State point 7 is consistent with state point 1 in terms of pressure. Based on the temperature and pressure data of each state point, the enthalpy, entropy, density, and other thermodynamic parameters corresponding to each state point are calculated using thermodynamic relationships. A working fluid state correlation table was established, containing 7 state points. Each state point includes parameters such as temperature, pressure, enthalpy, entropy, and density. Taking R245fa working fluid as an example, the temperature at evaporator inlet state point 1 is approximately 40℃ and the pressure is approximately 1.8MPa; the temperature at evaporator outlet state point 2 is approximately 130℃ and the pressure is approximately 4.5MPa; and the temperature at expander outlet state point 4 is approximately 80℃ and the pressure is approximately 0.3MPa.

[0036] A pressure-temperature pairing analysis was performed on the working fluid state correlation table to identify the phase response in the supercritical region. For each state point in the working fluid state correlation table, it was determined whether its pressure exceeded the critical pressure P_c and its temperature exceeded the critical temperature T_c. When both pressure and temperature exceeded the critical values, the state point was located in the supercritical region. Taking R245fa working fluid as an example, the critical pressure is 3.651 MPa and the critical temperature is 154.01℃. When the evaporator outlet pressure reaches 4.5 MPa and the temperature reaches 130℃, the working fluid is in the supercritical pressure region. Pressure-temperature pairing analysis was performed on the supercritical region state points to analyze the transition of the working fluid from the subcritical to the supercritical high-temperature region during evaporation, and the transition from the supercritical to the subcritical region during expansion. In the supercritical region, the working fluid exhibits liquid-like and gas-like phase transition characteristics. The isobaric specific heat capacity corresponding to each state point was determined, and the temperature and pressure corresponding to the peak position of the specific heat capacity were calculated. This position marks the transition from the liquid-like to the gas-like state. By comparing the pressure-temperature pairings at each state point with the peak position of the specific heat capacity, the positional relationship of each state point relative to the transition line is determined. This forms the phase response characteristics, which include the phase category identifier (liquid-like / gas-like) for each state point, the temperature and pressure deviation from the quasi-critical point, and the determination result of liquid-like or gas-like state characteristics. Under the working fluid R245fa at a pressure of 4.0 MPa, the peak specific heat capacity corresponds to a temperature of approximately 156℃. State points with temperatures below 156℃ are marked as "liquid-like," and those above 156℃ are marked as "gas-like." The phase response comprehensively describes the thermodynamic behavior characteristics of each state point in the supercritical region.

[0037] Phase transition boundary parameters are extracted from the phase response to generate an efficiency benchmark database. The temperature T_peak and pressure P_peak corresponding to the peak specific heat capacity are defined as the quasi-critical point. The operating conditions under which the working fluid in the evaporator crosses the quasi-critical point are identified. When the evaporator inlet temperature is lower than the quasi-critical temperature and the outlet temperature is higher than the quasi-critical temperature, the working fluid crosses the quasi-critical point in the evaporator, undergoing a transition from a liquid-like state to a gas-like state. Parameters such as the evaporator inlet pressure, outlet pressure, temperature difference, and heat transfer under the condition of crossing the quasi-critical point are recorded as phase transition boundary parameters. The supercriticality of the working fluid at the expander inlet is defined as Δh_sc = h3 - h_c, where h3 is the enthalpy at the expander inlet, h_c is the enthalpy at the critical point, and Δh_sc is the supercriticality. The larger the supercriticality, the deeper the working fluid penetrates into the supercritical region and the farther it is from the liquid-like-gas-like transition line. Taking R245fa as an example, its critical enthalpy is approximately 440 kJ / kg. When the expander inlet enthalpy is 480 kJ / kg, the supercriticality is 40 kJ / kg. System efficiency data under different supercriticalities were statistically analyzed to establish the correlation between supercriticality and system efficiency. The system efficiency reaches its peak when the supercriticality is within its optimal range. Actual operating data shows that with R245fa as the working fluid, the system efficiency can reach 8.5%-9.2% within the supercriticality range of 35-45 kJ / kg, which is higher than the efficiency level when the supercriticality deviates from this range. The phase transition limit parameters, supercriticality, and corresponding system efficiency values ​​were integrated into an efficiency benchmark database. The efficiency benchmark database contains multiple sets of operating condition data, each set including quasi-critical point parameters, supercriticality, evaporation temperature, evaporation pressure, condensation temperature, condensation pressure, and the corresponding benchmark efficiency value.

[0038] In some embodiments, the step of detecting current operating condition deviation and generating efficiency deviation metric based on the efficiency benchmark database includes: matching the current operating condition with the efficiency benchmark database to identify the type of operating condition deviation; performing offset vector parsing on the type of operating condition deviation to generate a direction classification; performing deviation propagation parsing on the direction classification to generate a deviation propagation distance; and forming an efficiency deviation metric based on the deviation propagation distance.

[0039] Deviation types are identified by matching the current operating conditions with an efficiency benchmark database. The database is searched for the benchmark operating condition whose parameters are closest to the current operating condition, using normalized Euclidean distance as the similarity metric. The normalized distance is calculated. Where i∈{1,2} represents the evaporation side and condensation side respectively, ΔTᵢ and ΔPᵢ are the temperature and pressure deviations between the current operating condition and the reference operating condition, T_ref=50℃, P_ref=2.0MPa. The reference operating condition with the smallest distance d is selected as the matching reference, and the parameter deviations are calculated and the operating condition deviation type is identified. Deviation judgment thresholds are set as follows: evaporation temperature ±3℃, evaporation pressure ±0.2MPa, condensation temperature ±2℃, condensation pressure ±0.1MPa, and supercriticality ±5kJ / kg. The deviation type is identified based on the direction and magnitude of the deviation; for example, an evaporation temperature deviation exceeding +3℃ is identified as an excessively high evaporation temperature type, and an evaporation pressure deviation below -0.2MPa is identified as an excessively low evaporation pressure type.

[0040] The deviation types of operating conditions are classified by offset vector analysis. Each deviation type is represented as an offset vector in the parameter space. A higher evaporation temperature corresponds to a positive offset in the evaporation temperature dimension, a lower evaporation pressure corresponds to a negative offset in the evaporation pressure dimension, a higher condensation temperature corresponds to a positive offset in the condensation temperature dimension, and a higher condensation pressure corresponds to a positive offset in the condensation pressure dimension. When multiple deviation types exist, the offset vectors of each type are superimposed to obtain a comprehensive offset vector. The components of the comprehensive offset vector in each dimension are analyzed to determine the dominant offset direction. The vector magnitudes of the evaporation-side parameters (evaporation temperature, evaporation pressure) and the condensation-side parameters (condensation temperature, condensation pressure) offsets are calculated. When the evaporation-side magnitude is greater than 1.5 times the condensation-side magnitude, it is classified as an evaporation-side deviation direction. When the condensation-side magnitude is greater than 1.5 times the evaporation-side magnitude, it is classified as a condensation-side deviation direction. When the ratio of the evaporation-side to condensation-side magnitudes is between 0.67 and 1.5, it is classified as a comprehensive deviation direction.

[0041] For example, the step of performing deviation propagation analysis on the direction classification to generate deviation propagation distance includes: determining a propagation path network by performing deviation diffusion path tracking based on the direction classification; extracting deviation decay gradients to form direction-sensitive weights based on the propagation path network; and using the direction-sensitive weights to quantize the propagation distance of the direction classification to form the deviation propagation distance.

[0042] Deviation propagation path network is determined by deviation diffusion path tracing based on directional classification. The propagation path network uses the evaporator, expander, condenser, and working fluid pump as nodes, with directed connections between nodes representing the direction and path of deviation propagation. For the evaporation-side deviation, a higher evaporation temperature leads to an increase in the working fluid temperature at the evaporator outlet. This temperature deviation propagates to the expander inlet, affecting the expander inlet temperature and the initial state of the expansion process, thus impacting the expander outlet temperature and output power. The increased expander outlet temperature propagates to the condenser inlet, increasing the condenser heat exchange load and affecting the condenser outlet temperature. The increased condenser outlet temperature propagates to the working fluid pump inlet, potentially reducing the working fluid density and affecting the pump's pressurization performance. The pump outlet pressure change propagates back to the evaporator inlet, forming a closed-loop feedback. The evaporation-side deviation forms a path sequence in the propagation path network: evaporator → expander → condenser → working fluid pump → evaporator. For the condensation-side deviation, a higher condensation temperature leads to an increase in the working fluid temperature at the condenser outlet. This temperature deviation propagates to the working fluid pump inlet, potentially causing vaporization of the working fluid within the pump, affecting pump operation and outlet pressure. The low pump outlet pressure propagates to the evaporator inlet, reducing the evaporation pressure and affecting the pressure boundary conditions of the evaporation process. This reduced evaporation pressure propagates to the expander inlet, altering the expander inlet operating conditions, which in turn affects the expander outlet parameters and propagates to the condenser inlet. The condensation-side deviation forms a path sequence in the propagation path network: condenser → working fluid pump → evaporator → expander → condenser. For the combined deviation direction, both evaporation-side and condensation-side deviations exist simultaneously, forming a cross-propagation path structure in the propagation path network. The propagation path network comprehensively describes the diffusion law of operating condition deviations in the circulation system.

[0043] The deviation attenuation gradient is extracted from the propagation path network to form direction-sensitive weights. As the deviation propagates between devices, its influence gradually decreases. The deviation influence intensity is defined as the relative rate of change of device performance parameters. At the evaporator, a 3°C increase in evaporation temperature leads to an increase of approximately 5 kJ / kg in outlet enthalpy, with an initial influence intensity of 1.0. When the deviation propagates to the expander, the 5 kJ / kg increase in expander inlet enthalpy leads to an increase of approximately 0.3 kW in output power, with a relative power change rate of approximately 3%, an attenuation coefficient of approximately 0.6, and an influence intensity decreasing to 0.6. As the deviation continues to propagate to the condenser, the increased condenser inlet temperature leads to a decrease in heat exchange temperature difference and a decline in condensation efficiency, with an attenuation coefficient of approximately 0.8 and an influence intensity decreasing to 0.48. When the deviation propagates to the working fluid pump, the temperature deviation has a relatively small impact on pump performance, with an attenuation coefficient of approximately 0.9 and an influence intensity decreasing to 0.43. The deviation attenuation gradient varies for different directional classifications. Deviations on the evaporator side attenuate more rapidly at the expander because the expander is sensitive to changes in inlet temperature; an attenuation coefficient of 0.6 is assigned to the expander. The deviation on the condenser side decays more rapidly at the working fluid pump because the pump is sensitive to changes in inlet conditions; a decay coefficient of 0.5 is assigned to the working fluid pump. Based on the differences in decay coefficients at different equipment for each direction category, a direction-sensitive weight is formed. For the evaporator-side deviation direction, the direction-sensitive weight is set to 1.5 for the expander, 1.0 for the condenser, and 0.8 for the working fluid pump. For the condenser-side deviation direction, the direction-sensitive weight is set to 1.5 for the working fluid pump, 1.2 for the evaporator, and 0.9 for the expander.

[0044] The deviation propagation distance is quantified by classifying directions using directional sensitivity weights. The deviation propagation distance is defined as the cumulative influence intensity of the deviation as it propagates from the source device to the target device. For deviations on the evaporator side, the initial influence intensity at the evaporator is set as the baseline value I_0. After propagating to the expander, the influence intensity is calculated based on the expander's directional sensitivity weight and attenuation coefficient. The change in influence intensity is ΔI_1 = I_0 × (1 - α_1) × w_1, where α_1 is the attenuation coefficient at the expander, w_1 is the expander's directional sensitivity weight, and the cumulative propagation distance is d_1 = ΔI_1. The deviation continues to propagate to the condenser and working fluid pump, with the propagation distance gradually accumulating: d_2 = d_1 + I_1 × (1 - α_2) × w_2, d_3 = d_2 + I_2 × (1 - α_3) × w_3. The deviation propagates back to the evaporator, forming a closed loop. The total propagation distance is d_total = d_3 + I_3 × (1 - α_4) × w_4, representing the cumulative decrease in influence intensity along the entire loop path. For the deviation direction on the condenser side, a similar method is used to calculate the total propagation distance from the condenser through the working fluid pump, evaporator, expander, and back to the condenser. For the combined deviation direction, the propagation distances of the evaporator-side deviation and the condenser-side deviation are calculated, and the weighted average of the two is taken as the combined propagation distance: d_total = β × d_evap + (1 - β) × d_cond, where β is the evaporator-side weighting coefficient, determined based on the modulus ratio of the evaporator-side and condenser-side deviation vectors.

[0045] An efficiency deviation metric is established based on the deviation propagation distance. A quantitative relationship between the deviation propagation distance and system efficiency deviation is established, and the system efficiency variation pattern under different deviation propagation distances is analyzed. Regression analysis is performed on actual operating data to fit the mapping function between propagation distance and efficiency deviation. The efficiency deviation metric is defined as Δη_measure = f(d_total), where Δη_measure is the efficiency deviation metric, f is the mapping function, and d_total is the deviation propagation distance. The mapping function adopts a quadratic polynomial form: Δη_measure = k_1 × d_total + k_2 × d_total², where k_1 and k_2 are fitting coefficients. For the supercritical organic Rankine cycle with R245fa working fluid, the fitting coefficients k_1 and k_2 are approximately 0.008 and 0.0015, respectively. The efficiency deviation metric value under the current operating condition is calculated. A positive efficiency deviation metric value indicates that the current efficiency is lower than the baseline efficiency; the larger the value, the more severe the efficiency deviation. When the efficiency deviation metric value exceeds 0.03 (i.e., 3%), it is marked as a significant deviation state, requiring the initiation of an operating condition adjustment strategy. When the deviation from the metric exceeds 0.05 (i.e., 5%), it is marked as a serious deviation and immediate intervention is required to restore normal operation.

[0046] Step S130: Perform load matching analysis between equipment to establish a cross-equipment response chain based on the cyclic operation characteristics. Determine the temporal coupling degree between the expander and the evaporator through the cross-equipment response chain. Use the temporal coupling degree to capture the load fluctuation response characteristics and extract the efficiency sensitive interval from the load fluctuation response characteristics.

[0047] Specifically, load matching analysis between equipment is conducted to establish a cross-equipment response chain based on the characteristics of cyclic operation. The cross-equipment response chain connects each piece of equipment according to the working fluid flow direction: evaporator → expander → condenser → working fluid pump → evaporator, describing the propagation path of load changes between equipment. The heat exchange load of the evaporator is determined by the heat source temperature, heat source flow rate, and working fluid flow rate. Changes in the heat exchange load directly affect the working fluid temperature and pressure at the evaporator outlet. Taking R245fa working fluid as an example, as the heat source flow rate increases, the evaporator heat exchange load increases accordingly, and the outlet temperature and pressure rise simultaneously, accelerating the transition of the working fluid from a subcritical to a supercritical state. The expander load is determined by the inlet working fluid state and rotational speed. Changes in the evaporator outlet state are transmitted to the expander inlet, affecting the expander load and output power. In the cross-equipment response chain, changes in expander output power affect the generator load, and changes in the expander outlet state are transmitted to the condenser inlet, affecting the condenser heat exchange load. Increased condenser heat exchange load leads to an increase in the outlet working fluid temperature, which is transmitted to the working fluid pump inlet, affecting the working fluid pump's operating status. An increase in the working fluid inlet temperature reduces the working fluid density, weakening the pump's pressurization effect at the same speed and causing a drop in outlet pressure. This change in pump outlet pressure is transmitted to the evaporator inlet, affecting the evaporator's inlet pressure boundary conditions and consequently influencing the pressure-temperature pairing relationship in the evaporation process, thus forming a closed-loop feedback mechanism in the cross-equipment response chain.

[0048] In some embodiments, determining the temporal coupling degree between the expander and the evaporator through the cross-device response chain includes: extracting the moment of sudden load change of the expander from the cross-device response chain as a trigger marker; monitoring the delay duration after the trigger marker when the evaporator temperature begins to respond; performing stability statistics on the delay duration to obtain a delay stability coefficient; and performing interval assignment determination on the delay stability coefficient to generate a temporal coupling degree.

[0049] The moment of load abrupt change in the expander is extracted from the cross-device response chain as a trigger marker. The expansioner load is monitored along the cross-device response chain, characterized by its output power. Time-series data of the expander output power is recorded at a sampling frequency of 1 Hz to ensure the capture of transient characteristics of load changes. The rate of change of the expander output power over time is obtained, reflecting the speed of load change. A moving average algorithm is used to calculate the power change rate, with a window width of 5 seconds to smooth instantaneous fluctuations. A load abrupt change judgment threshold is set; when the absolute value of the expander output power change rate exceeds this threshold, it is judged as a load abrupt change. The load abrupt change judgment threshold is determined based on the expander's rated power, typically set to 5% of the rated power per second. This threshold effectively identifies load jumps caused by changes in system operating conditions while avoiding misjudging normal operating fluctuations as abrupt changes. The moment of the expander load abrupt change is identified, and the timestamp of the abrupt change is recorded. Multiple load abrupt changes may occur within a complete operating cycle, each corresponding to operating condition disturbances such as heat source flow adjustment, cooling water temperature change, or working fluid flow fluctuation. All abrupt change moments are recorded to form a trigger marker sequence. Trigger markers indicate critical moments when the expander load changes significantly. These moments correspond to transition points in the system's operating conditions and serve as the starting point for analyzing the timing response relationships between devices.

[0050] The system monitors the delay time after a trigger marker is set, indicating when the evaporator temperature begins to respond. For each trigger moment in the trigger marker sequence, the change in evaporator outlet temperature is monitored. Before a sudden load change, the evaporator outlet temperature is relatively stable with small temperature fluctuations. After a sudden expansion load change, due to the flow delay of the working fluid in the pipes, the thermal inertia of the evaporator heat exchanger, and the delay in the phase transition of the working fluid, the evaporator outlet temperature does not respond immediately. The evaporator outlet temperature is continuously monitored, and when it begins to deviate from the steady state and shows a significant trend, it is determined that the evaporator temperature has begun to respond. The rate of change of the evaporator outlet temperature over time is obtained, and the same sliding difference algorithm as the rate of change of the expansion power is used to ensure timing consistency. A response determination condition is set: when the rate of change of the evaporator outlet temperature exceeds twice the standard deviation of the temperature fluctuation under steady state, it is determined that a response has begun. The threshold setting for the response determination condition is based on a statistical significance test; twice the standard deviation corresponds to a 95% confidence level, ensuring that the identified response is a true response rather than random fluctuation. Record the moment when the evaporator temperature begins to respond. The delay time is defined as: τ = t_response - t_trigger, where τ is the delay time, t_response is the response time, and t_trigger is the trigger time. For all trigger flags in the trigger flag sequence, obtain the corresponding delay time one by one to form a delay time sequence.

[0051] For example, the step of performing stability statistics on the delay duration to obtain a delay stability coefficient includes: performing dispersion detection on the delay duration to obtain the duration fluctuation amplitude; performing a concentration evaluation on the duration fluctuation amplitude to generate a fluctuation convergence index; performing stability inverse encoding on the fluctuation convergence index to generate a stability quantization value; and performing fluctuation suppression conversion on the stability quantization value to generate a delay stability coefficient.

[0052] The dispersion of the delay duration is detected to obtain the amplitude of duration fluctuation. The delay duration sequence contains multiple delay duration values, reflecting the time series response characteristics under different operating conditions and disturbances. The standard deviation of the delay duration sequence is obtained. As a classic measure of dispersion, the standard deviation quantitatively describes the degree to which the delay duration is dispersed around the mean. The standard deviation calculation uses Bessel correction, dividing by the degrees of freedom n-1 instead of the sample size n, to improve the estimation accuracy in small sample cases. The range of the delay duration sequence is obtained. The range is defined as the difference between the maximum and minimum values ​​in the sequence, reflecting the fluctuation range of the delay duration under the most extreme operating conditions. The range is sensitive to outliers and can capture system response anomalies under special operating conditions. Combining the standard deviation and range, the duration fluctuation range is defined as: A_delay = w1 × σ_τ + w2 × R_τ, where A_delay is the duration fluctuation range, σ_τ is the standard deviation of the delay duration, R_τ is the range of the delay duration, and w1 and w2 are weighting coefficients, typically set to w1 = 0.6 and w2 = 0.4, with the sum of the weighting coefficients being 1. The weighting allocation reflects the relative importance placed on average dispersion and extreme fluctuations; a larger weight for the standard deviation indicates a greater focus on overall stability, while a smaller weight for the range, though still retained, indicates a focus on extreme operating conditions.

[0053] A concentration assessment of duration fluctuations is performed to generate a fluctuation convergence index. The concentration characteristics of delay durations are analyzed based on the duration fluctuation amplitude A_delay. The coefficient of variation (COP) is extracted from the delay duration series, defined as the ratio of the standard deviation to the mean, eliminating the influence of dimensions and making delay durations at different average levels comparable. A smaller COP indicates less fluctuation in delay duration relative to its average level, and better consistency in the time series response. The proportion of data points P_center falling within ±1 standard deviation of the mean in the delay duration series is statistically analyzed. This proportion reflects the degree of clustering of delay durations towards the center value. According to the normal distribution theory, approximately 68% of normally distributed data falls within ±1 standard deviation. The degree to which the actual data's central clustering proportion deviates from this theoretical value reflects the centrality of the distribution. Combining the discrete characteristics of latency fluctuation amplitude with the centralized characteristics of latency duration, the volatility convergence index is defined as a weighted average of the inverse coefficient of variation and the proportion of central clustering: C_conv = w_c1 × (1 / CV) + w_c2 × P_center, where C_conv is the volatility convergence index, CV is the coefficient of variation, and w_c1 and w_c2 are weighting coefficients, typically set to w_c1 = 0.7 and w_c2 = 0.3. The relatively large weight of the inverse coefficient of variation reflects the emphasis on relative volatility, while the relatively small but not negligible weight of the proportion of central clustering reflects the focus on distribution pattern. The volatility convergence index comprehensively reflects the stability level of latency duration after evaluation based on volatility amplitude.

[0054] Stability quantification values ​​are generated by inverse stability encoding of the fluctuation convergence index. A larger fluctuation convergence index value corresponds to higher stability. To facilitate understanding and use, the fluctuation convergence index is normalized. The fluctuation convergence index is mapped to the 0-1 interval using the logistic function: S_norm=1 / (1+exp(-k×(C_conv-C_0))), where S_norm is the normalized stability quantification value, k is the steepness coefficient controlling the steepness of the curve (usually k=2), and C_0 is the center parameter determining the curve position based on experimental data. The logistic function exhibits an S-shaped curve characteristic, compressing the large variation range of the fluctuation convergence index into the 0-1 interval while maintaining a monotonically increasing relationship. The logistic function changes fastest near the center parameter and slows down as it moves away from the center parameter, which meets the practical needs of stability evaluation: when stability is moderate, small changes in the fluctuation convergence index cause large changes in the stability evaluation; when stability is extremely high or low, changes in the fluctuation convergence index have little impact on the stability evaluation. Normalization eliminates the dimensional influence of fluctuation convergence indices, making the stability of different operating conditions and time periods comparable, and providing a unified measurement standard for stability comparison analysis across operating conditions and time periods.

[0055] The stability quantification value is converted into a fluctuation suppression degree to generate a delay stability coefficient. To enhance the discernibility of stability differences, a nonlinear transformation is applied to the stability quantification value to amplify the differences in high-stability regions. The delay stability coefficient is defined as: K_stable = 1 - exp(-α × S_norm), where K_stable is the delay stability coefficient, α is the transformation coefficient controlling the transformation strength (usually α = 3), and S_norm is the stability quantification value. This nonlinear transformation uses a negative exponential function. When the stability quantification value is small, the delay stability coefficient increases slowly, corresponding to regions with poor stability; small stability improvements do not significantly improve system synergy. When the stability quantification value is large, the delay stability coefficient increases rapidly, corresponding to regions with good stability; small stability improvements can significantly improve system synergy. This asymmetric response characteristic conforms to the actual law that device synergy improves significantly with increasing stability. The selection of the transformation coefficient α is based on experimental data fitting; the larger α is, the stronger the transformation and the greater the amplification of stability differences. The value range of the delay stability coefficient is 0-1, and the coefficient value directly reflects the tightness of timing coupling between devices.

[0056] The delay stability coefficient is used to determine the interval classification to generate the time coupling degree. Based on the numerical range of the delay stability coefficient, the time coupling degree is divided into levels. Three intervals are defined: strong coupling, medium coupling, and weak coupling. The thresholds for interval division are determined based on system control requirements and historical operating data statistics. When the delay stability coefficient is greater than 0.7, it is determined to be in the strong coupling interval, corresponding to a high time coupling degree. Strong coupling indicates that the time response between the expander and evaporator is highly stable, the delay duration fluctuates very little, load changes can be transmitted between the devices with stable delay characteristics, and the system has good predictability of dynamic response. When the delay stability coefficient is in the range of 0.4-0.7, it is determined to be in the medium coupling interval, corresponding to a medium time coupling degree. Medium coupling indicates that the time response has a certain degree of stability, but the delay duration fluctuates to a moderate extent, the load transmission process is affected by operating condition fluctuations, and the predictability of the system's dynamic response is generally low. When the delay stability coefficient is less than 0.4, it is determined to be in the weak coupling interval, corresponding to a low time coupling degree. Weak coupling indicates unstable timing response, drastic fluctuations in delay duration, poor coordination between expander and evaporator, significant uncertainty in load changes during transmission between equipment, and difficulty in predicting the system's dynamic response.

[0057] The load fluctuation response characteristics are captured using temporal coupling degree. The amplitude characteristics of the evaporator temperature response curve are analyzed based on the degree of temporal coupling. Under strong coupling, the sudden load change from the expander propagates rapidly and stably to the evaporator, resulting in small fluctuations in the evaporator temperature response amplitude and good dynamic response performance. Under weak coupling, the propagation of the sudden load change from the expander to the evaporator is unstable, leading to large fluctuations in the evaporator temperature response amplitude and poor dynamic response performance. The time-series characteristics of the evaporator temperature response curve under different degrees of temporal coupling are analyzed. Under strong coupling, the response curve exhibits a sharp single peak and concentrated amplitude, while under weak coupling, the response curve exhibits multi-peak oscillations and dispersed amplitude. Load fluctuation response characteristics are thus formed, comprising the time-series data and amplitude characteristics of the evaporator temperature response curve, describing the dynamic response characteristics of the system under different degrees of temporal coupling.

[0058] In some embodiments, extracting the efficiency-sensitive interval from the load fluctuation response features includes: performing amplitude decomposition on the load fluctuation response features to obtain a peak envelope; performing threshold crossing detection from the peak envelope to identify sensitive fluctuation segments; determining the response intensity based on the sensitive fluctuation segments; and dividing the sensitive domain based on the response intensity to determine the efficiency-sensitive interval.

[0059] Amplitude decomposition is performed on the load fluctuation response characteristics to obtain the peak envelope. The load fluctuation response characteristics include time-series data of the evaporator temperature response curve, which contains amplitude and phase information. Amplitude decomposition is performed on the response curve to separate the envelope characteristics and extract the amplitude variation over time. The upper and lower envelopes of the response curve are extracted using Hilbert transform or the local extremum method. The Hilbert transform method, based on signal processing theory, extracts instantaneous amplitude through complex analytic signals and is suitable for non-stationary signals. The local extremum method identifies local maxima and minima of the curve and uses cubic spline interpolation to connect the extrema points to form the envelope. This method is simple and intuitive and suitable for curves with obvious oscillatory characteristics. The upper envelope connects all local maxima points of the response curve, and the lower envelope connects all local minima points. The average value of the upper and lower envelopes is obtained as the center line of the response curve, which reflects the low-frequency trend component of the response curve. The peak envelope is defined as the difference between the upper envelope and the center line: A_peak(t) = T_upper(t) - T_center(t), where A_peak(t) is the peak envelope, T_upper(t) is the upper envelope, and T_center(t) is the center line. The peak envelope eliminates the trend component of the response curve while retaining the high-frequency variation characteristics of the amplitude.

[0060] Sensitive fluctuation segments are identified by threshold crossing detection from the peak envelope. An amplitude threshold is set as the criterion for distinguishing between high-amplitude and low-amplitude segments. The threshold is determined based on the statistical characteristics of the peak envelope to ensure that the identified sensitive segments are statistically significant. The mean and standard deviation of the peak envelope are obtained; the mean reflects the overall amplitude level, and the standard deviation reflects the degree of amplitude fluctuation. The threshold is set to the mean plus one standard deviation, corresponding to the upper quartile of the peak envelope distribution. Segments exceeding this threshold account for approximately 15-20% of the total time. Each data point of the peak envelope is iterated, and it is determined whether the value exceeds the threshold. When the peak envelope value exceeds the threshold, this moment is marked as a sensitive moment, corresponding to a moment with a large load fluctuation response amplitude. Consecutive sensitive moments constitute sensitive fluctuation segments, and the duration of the segment reflects the persistence of the high-amplitude response. All sensitive fluctuation segments in the peak envelope are identified, and the start and end times of each segment are recorded to form a sensitive fluctuation segment sequence.

[0061] The response intensity is determined based on sensitive fluctuation segments. For each sensitive fluctuation segment, feature parameters are extracted from three dimensions: amplitude, time, and energy. The maximum value of the peak envelope within the segment is extracted; the maximum value represents the maximum response amplitude within the segment. A larger amplitude indicates a stronger instantaneous impact of load fluctuations on the system. The duration of the sensitive fluctuation segment is obtained; the duration is the time difference between the end and start of the segment. A longer duration indicates a longer period of high amplitude in the system, resulting in a greater cumulative impact. The integral area of ​​the peak envelope within the sensitive fluctuation segment is obtained; the integral area reflects the cumulative response energy within the segment. A larger integral area indicates a stronger overall disturbance to the system caused by load fluctuations during that period. All parameters are normalized to eliminate dimensional differences, converting different physical quantities into dimensionless indices. The maximum response amplitude is normalized to A_norm = A_max / A_ref, where A_ref is the reference amplitude and is taken as a threshold value. The duration is normalized as t_norm = Δt / t_ref, where t_ref is the average duration of the reference period within the sensitive segment. The integral area is normalized as S_norm = S / S_ref, where S_ref is the product of the reference area (threshold value) and the reference duration. The response intensity is defined as a weighted sum of the normalized parameters: I_response = w_a × A_norm + w_d × t_norm + w_s × S_norm, where I_response is the response intensity, and w_a, w_d, and w_s are weighting coefficients determined based on the relative importance of each parameter to efficiency; typically, w_a = 0.5, w_d = 0.3, and w_s = 0.2. The weighting coefficients are determined based on regression analysis of historical operating data. The amplitude weight is the largest, reflecting the most significant impact of instantaneous high amplitude on the equipment; the duration weight is the second largest, reflecting the cumulative effect of sustained high amplitude; and the integral area weight is the smallest but not negligible, reflecting the impact of overall energy disturbance.

[0062] Efficiency-sensitive intervals are determined by dividing the sensitive region based on response intensity. Sensitive fluctuation segments are classified according to the magnitude of the response intensity, identifying the most significantly sensitive segments that impact efficiency. A response intensity threshold is set to divide the sensitive fluctuation segments into high-sensitivity and low-sensitivity segments. The response intensity threshold is determined based on the system's safety margin and efficiency requirements. Threshold setting needs to balance system safety and economy; a threshold that is too low leads to excessive false alarms and increased control costs, while a threshold that is too high leads to missed alarms and increased operational risks. The response intensity threshold is typically set as the average response intensity plus one standard deviation. This threshold is based on the principle of statistical significance; segments exceeding the threshold are statistically significantly different from general segments. When the response intensity exceeds the threshold, it is determined to be a high-sensitivity segment, and the corresponding time interval is the efficiency-sensitive interval. The efficiency-sensitive interval represents the period when load fluctuations have the most significant impact on system efficiency. Within these intervals, system stability is poor, and efficiency is highly sensitive to operating condition disturbances; even small load fluctuations can lead to large efficiency deviations. The occurrence of efficiency-sensitive ranges is often related to critical transitions in system operating conditions. For example, when the working fluid phase is close to the quasi-critical point, the thermophysical properties change drastically; when the expander speed is close to the rated value, the flow stability decreases; and when the condensation temperature is close to the cooling water temperature, the heat exchange temperature difference is too small.

[0063] Step S140: Amplify the efficiency deviation measurement within the efficiency sensitive range to generate equipment collaborative features. Perform working fluid circulation path tracing on the equipment collaborative features to generate a flow path index. Perform specific enthalpy-specific entropy transfer analysis on the flow path index to generate unmeasured parameters. Capture heat transfer efficiency change features from the unmeasured parameters to obtain efficiency cascade data.

[0064] Specifically, within the efficiency-sensitive interval, the efficiency deviation metric is amplified to generate equipment coordination characteristics. Within this interval, the impact of load fluctuations on efficiency is amplified, placing the system in a high-risk state. Therefore, sensitivity adjustments to the efficiency deviation metric are needed to highlight operational risks during critical periods. When the current time is within the efficiency-sensitive interval, the amplified efficiency deviation metric Δη_amplified = β(t) × Δη_measure, where β(t) is the amplification factor reflecting efficiency sensitivity within the sensitive interval, and Δη_measure is the efficiency deviation metric generated in step S120. The amplification factor is determined based on the response intensity of the sensitive interval at that time. A higher response intensity indicates a more significant impact of that interval on efficiency, and a higher amplification factor results in a greater degree of amplification of the efficiency deviation metric. The amplification factor is typically proportional to the response intensity: β(t) = 1 + γ × I_response, where γ is a proportionality coefficient controlling the amplification intensity, usually taken as γ = 0.5. When the current time is not within the efficiency-sensitive interval, the amplification factor is 1, and the efficiency deviation metric remains unchanged. Combining the temporal coupling degree, load fluctuation response characteristics, and the efficiency-sensitive interval, equipment coordination characteristics are generated. Equipment coordination characteristics comprehensively describe the coordinated operation of various devices in a supercritical organic Rankine cycle system under load fluctuations, quantifying the dynamic coupling relationships and efficiency sensitivity among the devices. These characteristics include the temporal coupling degree level between the expander and evaporator, the average delay duration of the load fluctuation response, the standard deviation of the delay duration, the delay stability coefficient, the average response intensity, the start and end times of the efficiency-sensitive intervals, the number of sensitive intervals, and the proportion of the sensitive intervals to the total operating time. These characteristics provide data support for system optimization control and fault early warning, guiding the formulation of control strategies and parameter tuning.

[0065] The working fluid circulation path is traced using equipment cooperative features to generate a flow path index. This index describes the flow path characteristics and spatial distribution of the working fluid in a supercritical organic Rankine cycle system. The spatial resolution of the path tracing is determined based on the temporal coupling degree and efficiency-sensitive region information from the equipment cooperative features. A coarser spatial resolution can be used for flow path tracing in sections with high temporal coupling, while a finer spatial resolution is needed to capture state gradients in efficiency-sensitive regions corresponding to rapidly changing working fluid states. Along the working fluid flow direction, the circulation path is divided into multiple flow segments, based on equipment boundaries, phase transition regions, and efficiency-sensitive regions. The flow path index comprises seven main flow segments: the section from the working fluid pump outlet to the evaporator inlet (high-pressure subcooled liquid), the flow segment inside the evaporator (the transition from subcooled liquid to supercritical fluid, a key energy input segment), the section from the evaporator outlet to the expander inlet (supercritical or superheated state), the flow segment inside the expander (rapid expansion and work, the core energy conversion segment), the section from the expander outlet to the condenser inlet (low-pressure superheated steam or two-phase state), the flow segment inside the condenser (heat release and condensation), and the section from the condenser outlet to the working fluid pump inlet (low-temperature subcooled liquid). Each flow segment is numbered and labeled to form a complete flow path index.

[0066] In some embodiments, the step of generating unmeasured parameters by performing specific enthalpy-specific entropy transfer analysis on the flow path index includes: establishing a node distribution along the working fluid flow path based on the flow path index; applying energy conservation constraints at the node distribution to obtain enthalpy transfer relationships; combining the enthalpy transfer relationships with the pressure information at the node distribution to back-calculate the local superheat of the working fluid; and validating the local superheat of the working fluid to generate unmeasured parameters.

[0067] A node distribution along the working fluid flow path is established based on a flow path index. Calculation nodes for the working fluid state parameters are set along each flow segment defined by the flow path index. Node settings follow the principle of state gradient adaptation: densely distributed nodes in areas with drastic changes in state parameters, and sparsely distributed nodes in areas with gradual changes in state parameters. Main nodes are set at the inlet and outlet of each device, corresponding to the sensor placement locations. Intermediate nodes are set inside each flow segment, corresponding to locations not covered by sensors, where the working fluid state parameters need to be calculated. Node spacing is set based on the flow segment length and the working fluid state change gradient. Inside the evaporator, the working fluid transitions from a liquid-like to a gas-like state, resulting in a large state change gradient; the node spacing is set to one-tenth of the evaporator length, approximately 10 intermediate nodes for a 10-meter-long evaporator. Inside the expander, the working fluid expands rapidly, resulting in drastic state changes; the node spacing is set to one-fifth of the expander channel length, approximately 5 intermediate nodes for a typical radial expander. In pipeline sections, the working fluid state changes are relatively small; the node spacing can be set to half the pipeline length, one intermediate node for a 2-meter-long pipeline is sufficient. All nodes are numbered consecutively from node 1 to node N according to the working fluid flow sequence, thus establishing a node distribution table. The node numbering order is consistent with the working fluid flow direction, which facilitates the establishment of enthalpy transfer relationships and iterative calculations.

[0068] Energy conservation constraints are applied at the node distribution points to obtain the enthalpy transfer relationship. For adjacent nodes inside the evaporator, the working fluid absorbs heat, increasing its enthalpy. The enthalpy transfer relationship is: h_j+1 = h_j + q_evap, where h_j is the enthalpy of upstream node j, h_j+1 is the enthalpy of downstream node j+1, and q_evap is the specific heat absorbed by the working fluid between nodes. The specific heat q_evap is calculated based on the local heat transfer coefficient, heat transfer area, and logarithmic mean temperature difference. The heat transfer coefficient is determined using the correlation formula under the corresponding operating conditions. The working fluid in the evaporator inlet section is a subcooled liquid with a high heat transfer coefficient and a large specific heat. The working fluid in the evaporator outlet section is close to a gaseous state with a low heat transfer coefficient and a small specific heat. For adjacent nodes inside the expander, the working fluid expands and does work, decreasing its enthalpy. The enthalpy transfer relationship is: h_j+1 = h_j - w_exp, where w_exp is the specific work done by the working fluid between nodes. Specific work w_exp is calculated based on the polytropic index of the expansion process and the pressure ratio. For adjacent nodes inside the condenser, the working fluid releases heat, and the enthalpy decreases. The enthalpy transfer relationship is: h_j+1=h_j-q_cond, where q_cond is the specific heat released by the working fluid between nodes. For adjacent nodes in the pipeline section, assuming adiabatic flow and a small pressure drop, the enthalpy remains constant, and the enthalpy transfer relationship is: h_j+1=h_j. The enthalpy transfer relationship between all adjacent nodes is established based on the node distribution. Starting from the known enthalpy value at the measurement point, the enthalpy value of intermediate nodes is gradually calculated along the direction of working fluid flow using a forward calculation method.

[0069] The local superheat of the working fluid is calculated by combining the enthalpy transfer relationship with the pressure information at the node distribution. For intermediate nodes in the node distribution, the enthalpy h of that node is obtained through the enthalpy transfer relationship. The working fluid pressure P at this node is measured or estimated. For nodes inside the evaporator and condenser, the pressure can be estimated using linear interpolation based on the inlet and outlet pressures. Linear interpolation assumes that the pressure drop is uniformly distributed along the flow direction. For flow sections with high flow resistance, a nonlinear interpolation method is used, where the pressure drop follows a quadratic function distribution along the flow direction. Based on the enthalpy h and pressure P of the working fluid, the working fluid temperature T, saturation temperature T_sat, and entropy s corresponding to that node are obtained by consulting the working fluid thermodynamic property table or calling the equation of state. The thermodynamic property table uses a bilinear interpolation method, interpolating the temperature, saturation temperature, and entropy value from a discrete table based on the enthalpy and pressure. The local superheat of the working fluid is defined as: ΔT_sh = T - T_sat, where ΔT_sh is the local superheat, T is the actual working fluid temperature, and T_sat is the saturation temperature at that pressure. When the working fluid is in the supercritical region, there is no clear concept of saturation temperature. Local superheat is defined as the difference between the working fluid temperature and the quasi-critical temperature, where the quasi-critical temperature is taken as the temperature corresponding to the peak specific heat capacity at that pressure. For all intermediate nodes in the nodal distribution, the local superheat and entropy of the working fluid are calculated inversely to form the local superheat distribution and entropy distribution.

[0070] The effectiveness of the local superheat of the working fluid is verified to generate parameters that are not directly measured. The calculated local superheat of the working fluid is then checked to ensure it meets physical constraints and the actual operating conditions of the cycle. For the evaporator outlet node, the working fluid should be in a supercritical or superheated steam state, and the local superheat should be positive and greater than the minimum superheat requirement, typically set at 5°C to prevent wet steam from entering the expander. For the condenser outlet node, the working fluid should be in a subcooled liquid state, and the local superheat should be negative with an absolute value greater than the minimum subcooling requirement, typically set at 3°C ​​to prevent cavitation in the working fluid pump. For the expander inlet node, the working fluid temperature should be higher than the evaporation temperature, and the local superheat should be positive to ensure no condensation occurs during expansion. The error accumulation in the enthalpy transfer relationship calculation process is checked by comparing the calculated enthalpy at the measurement node with the actual enthalpy obtained by measuring temperature and pressure. The relative error should be within the allowable range, typically less than 3%. For nodes that do not meet the validity criteria, adjust the estimated heat transfer or work in the enthalpy transfer relationship, correct the heat transfer coefficient or polytropic index, and recalculate until the validity criteria are met. The local superheat of the working fluid that passes the validity verification is considered as a non-directly measured parameter. Non-directly measured parameters include state parameters such as enthalpy, temperature, superheat, and entropy of all intermediate nodes in the node distribution, forming a complete distribution of the working fluid state parameters along the circulation path.

[0071] Efficiency gradient data is obtained by capturing the characteristics of heat transfer efficiency changes from unmeasured parameters. The unmeasured parameters of internal nodes in the evaporator and condenser are analyzed to identify the characteristics and spatial distribution of heat transfer efficiency changes. The working fluid temperature is low at the evaporator inlet, with a large temperature difference from the heat source, resulting in strong heat transfer driving force and high heat transfer efficiency. Conversely, the working fluid temperature is high at the evaporator outlet, with a small temperature difference from the heat source, resulting in weak heat transfer driving force and low heat transfer efficiency. The local heat transfer efficiency is defined as the ratio of the absorbed heat flux density to the maximum possible heat flux density at that node: η_local = q_actual / q_max, where q_actual is the actual heat flux density and q_max is the theoretical heat flux density when the temperature difference between the working fluid and the heat source reaches its maximum. The local heat transfer efficiency at each node within the evaporator is obtained, and the decay trend of heat transfer efficiency along the flow direction is identified. The local heat transfer efficiency in the first half of the evaporator is typically in the range of 70-90%, while the local heat transfer efficiency in the second half drops to the range of 30-50%. The local heat exchange efficiencies within the evaporator and condenser are classified into three levels: high efficiency, medium efficiency, and low efficiency. The high efficiency level corresponds to the node region with a local heat exchange efficiency greater than 70%, indicating good heat exchange performance. The medium efficiency level corresponds to the node region with a local heat exchange efficiency between 40% and 70%, indicating moderate heat exchange performance. The low efficiency level corresponds to the node region with a local heat exchange efficiency less than 40%, indicating poor heat exchange performance. The distribution length and proportion of each efficiency level in the evaporator and condenser are statistically analyzed to understand the spatial distribution characteristics of the efficiency levels. Efficiency level data provides a basis for optimized heat exchanger design and operation; adjusting the heat exchanger length distribution or working fluid flow distribution can improve the efficiency level distribution.

[0072] Step S150: Based on the efficiency deviation measurement and equipment coordination characteristics, predict the efficiency jump to determine the adjustment response time difference, and generate a graded adjustment command based on the adjustment response time difference and efficiency cascade data.

[0073] In some embodiments, the step of predicting efficiency jumps and determining the adjustment response time difference based on the efficiency deviation metric and the device coordination characteristics includes: extracting the efficiency dynamic rate by performing time-series change rate extraction on the efficiency deviation metric; performing correlation analysis on the efficiency dynamic rate and the device coordination characteristics to identify rate inflection points; performing jump risk assessment on the rate inflection points to generate an efficiency jump warning indicator; and determining the adjustment response time difference based on the interval between the efficiency jump warning indicator and the current time.

[0074] The efficiency deviation metric is extracted using a time-series rate of change to obtain the efficiency dynamic rate. The efficiency deviation metric changes over time, forming a time series, and the efficiency deviation metric values ​​at different times are recorded. The forward differencing method is used to calculate the time-series rate of change of the efficiency deviation metric: efficiency dynamic rate v(t) = [Δη(t) - Δη(t - Δt)] / Δt, where v(t) is the efficiency dynamic rate at time t, Δη(t) and Δη(t - Δt) are the efficiency deviation metrics at the current and previous times, and Δt is the time interval (sampling period). The time series of the efficiency dynamic rate is obtained, and the efficiency dynamic rate values ​​at each time point are recorded. The sign and amplitude changes of the efficiency dynamic rate reflect the evolution trend of system efficiency. When the efficiency dynamic rate is consistently positive and its value increases, it indicates that efficiency is deteriorating rapidly, with a risk of abrupt change. When the efficiency dynamic rate changes from negative to positive, it indicates that efficiency is shifting from improvement to deterioration, which may be an early signal of operating condition deviation. The efficiency dynamic rate is filtered using a moving average filter to suppress rate fluctuations caused by measurement noise, with the filter window width set to 3-5 sampling periods.

[0075] Correlation analysis is performed on the efficiency dynamic rate and equipment coordination characteristics to identify rate inflection points. Equipment coordination characteristics include information such as temporal coupling degree, efficiency sensitive interval, and response intensity, providing a criterion for rate inflection point identification. The correlation between efficiency dynamic rate and equipment coordination characteristics is analyzed to identify moments when the efficiency dynamic rate undergoes significant changes. A rate inflection point is defined as the moment when the rate of change of efficiency dynamic rate changes sign or abruptly changes amplitude. The rate of change of efficiency dynamic rate, i.e., the derivative of efficiency dynamic rate with respect to time, reflects the acceleration of efficiency change. The central difference method is used to calculate the rate of change of dynamic rate, which has higher accuracy than forward difference. When the rate of change of efficiency dynamic rate changes from negative to positive, it indicates that the rate of efficiency deterioration has shifted from slowing down to accelerating, corresponding to a rate inflection point. When the amplitude of the rate of change of efficiency dynamic rate suddenly increases by more than twice the standard deviation of normal fluctuations, it indicates that efficiency change has entered an acceleration phase, also corresponding to a rate inflection point. Combining the efficiency sensitive interval information of equipment coordination characteristics, it is determined whether the rate inflection point is located within the sensitive interval. Rate inflection points located within the efficiency-sensitive range carry a higher risk level because the impact of load fluctuations on efficiency is amplified within this range, meaning even small deviations in operating conditions can lead to significant efficiency jumps. Analyzing the relationship between rate inflection points and time-series coupling reveals that when time-series coupling is low, efficiency changes at rate inflection points are more unpredictable and more difficult to regulate. All rate inflection points meeting the specified conditions are identified, and their timing, dynamic efficiency rate value, rate of change amplitude, and the efficiency-sensitive range they fall within are recorded.

[0076] A risk assessment of efficiency jumps is performed at rate inflection points to generate an early warning indicator. For each identified rate inflection point, the risk level of an efficiency jump at that point is evaluated. The jump risk assessment comprehensively considers factors such as the amplitude of the efficiency dynamic rate, the amplitude of the rate of change, the response intensity of the corresponding efficiency-sensitive interval, and the temporal coupling degree. The jump risk index is defined as a comprehensive score of each factor, and the scoring method adopts fuzzy comprehensive evaluation, where each factor is normalized and weighted before being summed. The efficiency dynamic rate is normalized as the ratio of the current rate to the historical maximum rate, reflecting the relative drasticness of the current efficiency change, with a weight set at 0.4. The efficiency dynamic rate of change is normalized as the ratio of the current rate of change to the historical maximum rate of change, reflecting the relative accelerating trend of efficiency change, with a weight set at 0.3. The response intensity of the corresponding sensitive interval is directly used as the normalized sensitivity index, with a weight set at 0.2. The reciprocal of the temporal coupling degree, after normalization, reflects the additional risk caused by poor equipment coordination, with a weight set at 0.1. The jump risk index ranges from 0 to 1, with higher values ​​indicating higher jump risk. Jump risk thresholds are set: 0.7 for high risk and 0.5 for medium risk, categorizing rate inflection points into three levels: high, medium, and low risk. When the jump risk index exceeds 0.7, a high-risk early warning sign is generated. When the jump risk index is between 0.5 and 0.7, a medium-risk early warning sign is generated. When the jump risk index is below 0.5, a low-risk early warning sign is generated, or no warning is generated. The generation time, risk level, and expected jump time of the efficiency jump early warning sign are recorded. The expected jump time is calculated based on the efficiency dynamic rate and rate of change, assuming the efficiency dynamic rate continues to increase at the current acceleration, estimating the time when the efficiency deviation metric reaches the jump critical value.

[0077] The adjustment response time difference is determined based on the interval between the efficiency jump warning indicator and the current time. For the generated efficiency jump warning indicator, the time interval between the expected jump occurrence time and the current time is obtained. This interval is the warning time window, representing the available adjustment time from the current time to the expected efficiency jump occurrence. The adjustment response time difference consists of three parts: command transmission delay, actuator response time, and system state adjustment time. The command transmission delay depends on the communication speed and data processing time of the control system, and is typically 1-2 seconds for modern distributed control systems. The actuator response time depends on the operating speed of equipment such as valves, pumps, and frequency converters. The response time of electric valves is typically 5-10 seconds, the response time of frequency converters is typically 2-5 seconds, and the response time of working fluid pump speed regulation is typically 3-8 seconds. The system state adjustment time is determined based on the temporal coupling degree. When the temporal coupling degree is high, load changes are transmitted quickly between equipment, resulting in a short state adjustment time; when the temporal coupling degree is low, the transmission is slow, resulting in a long adjustment time. The system state adjustment time is determined based on the timing coupling level. When the timing coupling is high, the adjustment time is the average of the delay durations. When the timing coupling is medium, the adjustment time is the average of the delay durations plus one standard deviation. When the timing coupling is low, the adjustment time is the average of the delay durations plus two standard deviations. The adjustment response time difference is obtained by adding the instruction transmission delay, the actuator response time, and the system state adjustment time.

[0078] A tiered control command is generated based on the control response time difference and efficiency gradient data. The issuance time of the control command is determined based on the control response time difference. A lead time parameter is defined as the time preceding the issuance of the control command relative to the expected occurrence of the efficiency jump. The lead time parameter is t_advance = τ_response + t_buffer, where τ_response is the control response time difference and t_buffer is the buffer time, typically set to 20%-30% of the control response time difference. The issuance time of the control command, determined by the lead time parameter, is the expected occurrence of the efficiency jump minus the lead time parameter, ensuring that the control measures take effect before the efficiency jump occurs. The control target, direction, and magnitude are determined by integrating efficiency gradient data. The efficiency gradient data includes the proportion of high-efficiency, medium-efficiency, and low-efficiency stages in the evaporator and condenser. When the proportion of low-efficiency stages exceeds 40%, it indicates a severe degradation in heat exchanger performance, generating a large-scale control command with a magnitude of 10-20% of the rated value. The control targets include increasing the heat source flow rate or the working fluid flow rate. When the proportion of low-efficiency stages is between 20% and 40%, it indicates a moderate decline in heat exchanger performance, generating a moderate adjustment command with an adjustment range of 5-10% of the rated value, allowing for appropriate adjustments to operating parameters. When the proportion of low-efficiency stages is below 20%, it indicates good heat exchanger performance, generating a small adjustment command with an adjustment range of 2-5% of the rated value, allowing for fine-tuning or monitoring only. The graded adjustment command includes information such as the adjustment object, adjustment direction, adjustment range, and execution time. The adjustment object is determined based on the weakest link in the efficiency cascade data, which can be heat source flow rate, cooling water flow rate, working fluid pump speed, or expander load. The adjustment direction is determined by the sign of the efficiency deviation metric; a positive deviation indicates that efficiency is below the baseline, requiring enhanced heat exchange or reduced load; a negative deviation indicates that efficiency is above the baseline, allowing for appropriate reduction of operating costs.

[0079] To implement the supercritical organic Rankine cycle efficiency optimization method and system corresponding to the above method embodiments, in order to achieve the corresponding functions and technical effects. See also Figure 2 , Figure 2 A structural block diagram of the supercritical organic Rankine cycle efficiency optimization method and system 200 provided in this embodiment is shown. For ease of explanation, only the parts relevant to this embodiment are shown. The supercritical organic Rankine cycle efficiency optimization system 200 provided in this embodiment includes:

[0080] Data acquisition module 201 is used to acquire multi-device monitoring data during the operation of supercritical organic Rankine cycle system, and to perform operating condition offset tracing processing on the multi-device monitoring data to generate cycle operation characteristics;

[0081] Efficiency evaluation module 202 is used to establish a working fluid state association table based on the cyclic operation characteristics, perform supercritical region pressure-temperature pairing identification phase response on the working fluid state association table, extract phase transition limit parameters from the phase response to generate an efficiency benchmark database, and perform current operating condition deviation detection based on the efficiency benchmark database to generate an efficiency deviation metric.

[0082] The collaborative analysis module 203 is used to perform load matching analysis between equipment to establish a cross-equipment response chain for the cyclic operation characteristics, determine the time coupling degree between the expander and the evaporator through the cross-equipment response chain, capture the load fluctuation response characteristics using the time coupling degree, and extract the efficiency sensitive interval from the load fluctuation response characteristics.

[0083] The parameter generation module 204 is used to amplify the efficiency deviation measurement within the efficiency sensitive range to generate equipment collaborative features, perform working fluid circulation path tracking on the equipment collaborative features to generate a flow path index, perform specific enthalpy-specific entropy transfer analysis on the flow path index to generate unmeasured parameters, and capture heat transfer efficiency change features from the unmeasured parameters to obtain efficiency cascade data.

[0084] The instruction generation module 205 is used to predict efficiency jumps based on the efficiency deviation metric and the equipment coordination characteristics to determine the adjustment response time difference, and generate graded adjustment instructions based on the adjustment response time difference and the efficiency tier data.

[0085] The supercritical organic Rankine cycle efficiency optimization method and system 200 described above can implement the supercritical organic Rankine cycle efficiency optimization method and system of the above method embodiments. The options in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining content of this application embodiment can be referred to the content of the above method embodiments, and will not be repeated in this embodiment.

[0086] The purpose of the above embodiments is to reproduce and derive the technical solution of the present invention by way of example, and to fully describe the technical solution, purpose and effect of the present invention. The purpose is to enable the public to have a more thorough and comprehensive understanding of the disclosure of the present invention, and not to limit the scope of protection of the present invention.

[0087] The above embodiments are not an exhaustive list based on the present invention, and there may be many other embodiments not listed. Any substitutions and improvements made without departing from the concept of the present invention are within the protection scope of the present invention.

Claims

1. A method for optimizing the efficiency of a supercritical organic Rankine cycle, characterized in that, include: Multi-device monitoring data of a supercritical organic Rankine cycle system is collected during operation. The system undergoes condition offset tracing processing to generate cycle operation characteristics. This includes: scanning the multi-device monitoring data for efficiency fluctuations to obtain a fluctuation amplitude sequence; locating a steady-state benchmark segment from the fluctuation amplitude sequence using a benchmark value; using the steady-state benchmark segment as the starting point for tracing backwards to pinpoint the location of the first efficiency drop; and extracting equipment state parameters based on the location of the first efficiency drop to construct cycle operation characteristics. A working fluid state association table is established based on the cyclic operation characteristics. The working fluid state association table is used to identify the phase response by supercritical pressure-temperature pairing. The phase transition limit parameter is extracted from the phase response to generate an efficiency benchmark database. Based on the efficiency benchmark database, the current operating condition deviation is detected to generate an efficiency deviation metric. Based on the cyclic operation characteristics, load matching analysis between equipment is performed to establish a cross-equipment response chain. The time-series coupling degree between the expander and the evaporator is determined through the cross-equipment response chain. The time-series coupling degree is used to capture load fluctuation response characteristics, and the efficiency-sensitive interval is extracted from the load fluctuation response characteristics. Within the efficiency-sensitive range, the efficiency deviation measurement is amplified to generate equipment collaborative features. The equipment collaborative features are used to perform working fluid circulation path tracking to generate a flow path index. The flow path index is analyzed by specific enthalpy-specific entropy transfer to generate unmeasured parameters. The heat transfer efficiency change features are captured from the unmeasured parameters to obtain efficiency cascade data. The efficiency deviation metric and the equipment coordination characteristics are used to predict efficiency jumps and determine the adjustment response time difference. Based on the adjustment response time difference and the efficiency tier data, a graded adjustment command is generated.

2. The method according to claim 1, characterized in that, The step of generating an efficiency deviation metric based on the efficiency benchmark database for current operating condition deviation detection includes: Based on the current operating conditions and the efficiency benchmark database, deviation matching is performed to identify the type of operating condition deviation; The deviation types of the operating conditions are analyzed by offset vector parsing to generate direction classification; The deviation propagation distance is generated by performing deviation propagation analysis on the directional classification. The deviation in propagation distance forms an efficiency deviation metric.

3. The method according to claim 1, characterized in that, The determination of the temporal coupling degree between the expander and the evaporator through the cross-device response chain includes: Extract the moment of sudden change in expander load from the cross-device response chain as a trigger marker; Monitor the delay time after the trigger flag is set, during which the evaporator temperature begins to respond. The delay duration is statistically analyzed to obtain the delay stability coefficient; The time-series coupling degree is generated by determining the interval affiliation of the delay stability coefficient.

4. The method according to claim 1, characterized in that, The step of generating unmeasured parameters by performing enthalpy-entropy transfer analysis on the flow path index includes: The node distribution along the working fluid flow path is established based on the flow path index; Energy conservation constraints are applied at the node distribution locations to obtain the enthalpy transfer relationship; The local superheat of the working fluid is calculated by combining the enthalpy transfer relationship with the pressure information at the node distribution. The effectiveness of the local superheat of the working fluid is verified to generate parameters that are not directly measured.

5. The method according to claim 1, characterized in that, The step of determining the adjustment response time difference by predicting efficiency jumps based on the efficiency deviation metric and the equipment collaborative characteristics includes: The efficiency deviation metric is used to extract the time-series rate of change to obtain the efficiency dynamic rate. A correlation analysis was performed between the efficiency dynamic rate and the device collaborative characteristics to identify rate inflection points. A risk assessment of the rate inflection point is performed to generate an efficiency jump warning indicator. The adjustment response time difference is determined based on the interval between the efficiency jump warning flag and the current time.

6. The method according to claim 1, characterized in that, The extraction of the efficiency-sensitive interval from the load fluctuation response characteristics includes: The peak envelope is obtained by amplitude decomposition of the load fluctuation response characteristics; Sensitive fluctuation sections are identified by performing threshold crossing detection from the peak envelope; The response intensity is determined based on the aforementioned sensitive fluctuation range; The efficiency-sensitive interval is determined by dividing the sensitive domain based on the response intensity.

7. The method according to claim 2, characterized in that, The step of performing deviation propagation analysis on the direction classification to generate deviation propagation distance includes: Based on the aforementioned directional classification, deviation diffusion path tracking is performed to determine the propagation path network; Based on the propagation path network, deviation decay gradients are extracted to form direction-sensitive weights; The direction-sensitive weights are used to quantify the propagation distance of the direction classification to form the deviation propagation distance.

8. The method according to claim 3, characterized in that, The step of performing stability statistics on the delay duration to obtain the delay stability coefficient includes: The dispersion of the delay duration is detected to obtain the duration fluctuation amplitude; The fluctuation amplitude of the duration is centrally evaluated to generate a fluctuation convergence index. The stability quantization value is generated by performing stability inverse encoding on the fluctuation convergence index. The stability quantification value is converted into a fluctuation suppression degree to generate a delay stability coefficient.

9. A supercritical organic Rankine cycle efficiency optimization system, characterized in that, include: The data acquisition module is used to collect multi-device monitoring data during the operation of a supercritical organic Rankine cycle system, and to perform condition offset tracing processing on the multi-device monitoring data to generate cycle operation characteristics. This includes: scanning the multi-device monitoring data for efficiency fluctuations to obtain a fluctuation amplitude sequence; locating a steady-state reference segment from the fluctuation amplitude sequence using a benchmark value; using the steady-state reference segment as the tracing starting point to search backwards and locate the first efficiency drop position; and extracting equipment state parameters based on the first efficiency drop position to construct cycle operation characteristics. The efficiency evaluation module is used to establish a working fluid state association table based on the cyclic operation characteristics, perform supercritical pressure-temperature pairing identification of the working fluid state association table to identify the phase response, extract phase transition limit parameters from the phase response to generate an efficiency benchmark database, and perform current operating condition deviation detection based on the efficiency benchmark database to generate an efficiency deviation metric. The collaborative analysis module is used to perform load matching analysis between equipment to establish a cross-equipment response chain for the cyclic operation characteristics, determine the time coupling degree between the expander and the evaporator through the cross-equipment response chain, capture the load fluctuation response characteristics using the time coupling degree, and extract the efficiency sensitive interval from the load fluctuation response characteristics. The parameter generation module is used to amplify the efficiency deviation measurement within the efficiency sensitive range to generate equipment collaborative features, perform working fluid circulation path tracing on the equipment collaborative features to generate a flow path index, perform specific enthalpy-specific entropy transfer analysis on the flow path index to generate unmeasured parameters, and capture heat transfer efficiency change features from the unmeasured parameters to obtain efficiency cascade data. The instruction generation module is used to predict efficiency jumps based on the efficiency deviation metric and the equipment coordination characteristics to determine the adjustment response time difference, and generate graded adjustment instructions based on the adjustment response time difference and the efficiency tier data.

Images