A port green intelligent heat supply system based on a distributed air source heat pump

Abstract

This invention discloses a port green intelligent heating system based on a distributed air source heat pump, mainly involving the field of port green heating technology. It addresses the shortcomings of traditional heat pump control, which relies entirely on the lagging sampling data of temperature sensing bulbs. This makes it unable to match the dynamic changes in the thermal boundary of the port's wind field, and easily leads to distortion of thermal state perception and severe phase lag. This invention is operated in a unified manner by a wind field rhythm sensing sub-center. Through high-frequency wind speed sampling, multi-dimensional wind field feature quantification, dynamic assessment of thermal reliability, extrapolation correction of superheat trend, adaptive modulation of EEV action beat, and closed-loop iterative calibration of control parameters, the system uses a hierarchical control logic to accurately identify complex wind field conditions such as port laminar flow, high-frequency gusts, slowly changing sea and land winds, and impact / suction type asymmetric gusts. It dynamically switches the superheat control benchmark, adaptively adjusts the valve adjustment beat and asymmetric adjustment step size, and continuously optimizes the control parameters in a closed loop.

Description

Technical Field

[0001] This invention relates to the field of green heating technology for ports, specifically a green and intelligent heating system for ports based on a distributed air source heat pump. Background Technology

[0002] Against the backdrop of the port's green and intelligent transformation, traditional port heating modes using coal, gas, and electricity have many problems, such as high energy consumption, large carbon emissions, high operation and maintenance costs, and excessive pollution emissions. In contrast, distributed air source heat pumps, with their advantages of being clean and low-carbon, having controllable energy consumption, and flexible layout, have gradually become the mainstream green heating solution for port building heating, equipment heat tracing, and process heating, and are the core technology direction for zero-carbon transformation of port thermal systems.

[0003] However, the current mainstream port air source heat pump heating systems all use the fixed logic control architecture of traditional civil and commercial heat pumps. The control core relies entirely on the temperature sampling data of the evaporator outlet temperature sensor to complete the electronic expansion valve opening adjustment and superheat closed-loop control. It is only suitable for normal indoor, windless, and stable working conditions with stable wind speed. It has not been specifically optimized for the special open-air working environment of ports, which has led to the continuous amplification of the defects in the operation of heat pump systems in port scenarios.

[0004] In existing technologies, traditional heat pump control relies entirely on the lag sampling data of the temperature sensing bulb, which cannot match the changes in the thermal boundary of the dynamic wind field in the port. The thermal state perception is distorted and the phase lag is severe. Furthermore, the fixed EEV control cycle is prone to phase resonance with the rhythm of the port wind field disturbance, resulting in drastic fluctuations in superheat and a decrease in system energy efficiency.

[0005] To address the aforementioned technical shortcomings, a green and intelligent port heating system based on a distributed air source heat pump is proposed, aiming to provide an efficient, stable, and intelligent technical solution for zero-carbon intelligent heating in ports. Summary of the Invention

[0006] The purpose of this invention is to solve the problems mentioned above by proposing a green and intelligent heating system for ports based on distributed air source heat pumps.

[0007] The objective of this invention can be achieved through the following technical solution: a port green and intelligent heating system based on a distributed air source heat pump, including a wind field rhythm sensing sub-center, which includes a feature recognition unit, an effective reference construction unit, a controller action parameter self-modulation unit, and a closed-loop calibration dynamic adjustment unit.

[0008] The feature recognition unit is used to collect the instantaneous wind speed sequence of the air inlet surface of the heat pump unit at a preset sampling frequency, slice the sequence into sliding windows according to a preset window length, and extract the following three features in each window: the disorder index of wind speed fluctuation, the energy concentration frequency band of wind speed fluctuation, and the distribution skew direction of wind speed change rate.

[0009] The effective benchmark construction unit is used to dynamically evaluate the reliability of the current temperature sensing bulb reading based on the disorder index and energy concentration frequency band, and to weight and fuse the current measured superheat value with the trend extrapolation value to obtain the effective superheat deviation.

[0010] The controller action parameter self-modulation unit is used to calculate the wind field disturbance period based on the energy concentration frequency band, and calculate the active dormancy window of the electronic expansion valve based on the preset proportional coefficient k. At the same time, it dynamically corrects the direction step multiplier of the electronic expansion valve opening adjustment according to the sign of the distribution skew direction.

[0011] The closed-loop calibration dynamic adjustment unit is used to monitor the superheat response curve after each action of the electronic expansion valve, extract the time when the superheat response extreme value occurs, dynamically correct the k value based on the comparison result between the time and the current active sleep window, and feed back the corrected k value to the controller action parameter self-modulation unit.

[0012] Furthermore, the wind speed sequence within the window is arranged in chronological order to form a phase space, the embedding dimension is set to a first preset value, the similarity tolerance is set to a second preset multiple of the standard deviation of wind speed within the window, and the mean of the log-conditional probability of the new state generated under this tolerance is calculated and taken as a negative value.

[0013] Furthermore, the method for extracting the energy concentration frequency band is as follows: perform a discrete Fourier transform on the wind speed sequence within the window, calculate the power spectral density, and extract the frequency value corresponding to the peak value of the power spectral density.

[0014] Furthermore, the first derivative of wind speed with respect to time at each moment within the window is calculated, and then the ratio of the third central moment of the first derivative sequence to the cube of the standard deviation is calculated. A positive value of this ratio indicates a sudden rise and fall of wind speed in an impact gust, a negative value indicates a slow rise and fall of wind speed in a suction gust, and a value approaching zero indicates symmetrical wind speed fluctuations.

[0015] Furthermore, in the effective benchmark construction unit, the thermal reliability weight W of the current temperature sensor reading is... trust According to the disorder index H v The thermal inertia time constant τ of the temperature sensing bulb sensor The ratio of the preset window length L to the given value is calculated as follows:

[0016] , where λ is the preset scaling coefficient;

[0017] The trend extrapolation value SH pred (t) Calculated based on the current measured superheat value, the rate of change of the measured superheat value, and the rate of change of wind speed, using the following formula:

[0018] Where Δt is the sampling interval and β is the preset coupling coefficient;

[0019] The effective superheat deviation is calculated using the following formula:

[0020] , of which SH target Set the target value for superheat.

[0021] Furthermore, in the controller action parameters from the modulation unit, the active sleep window T dwell =k×T disturb T disturb =1 / f p f p This refers to the energy concentration frequency band;

[0022] When the distribution skew direction is greater than the first direction threshold, the valve opening direction step size multiplier is set as the first multiplier and the valve closing direction step size multiplier is set as the second multiplier.

[0023] When the distribution skew direction is less than the second direction threshold, the valve closing direction step size multiplier is set as the first multiplier and the valve opening direction step size multiplier is set as the second multiplier.

[0024] When the distribution skew direction is between the second direction threshold and the first direction threshold, the valve opening and closing step multipliers are both kept at 1, and the effective opening adjustment step of the electronic expansion valve is determined by the basic step and the corresponding direction multiplier.

[0025] Furthermore, in the closed-loop calibration dynamic adjustment unit, the time T at which the extreme value of the response occurs is compared with the current T. dwell Comparison: If T ≥ first proportional threshold × T dwell Then, the value of k is reduced by the first adjustment step; if T ≤ the second proportional threshold × T dwell If T is between the second proportional threshold and the first proportional threshold, then the k value is increased by the first adjustment step size; if T is between the second proportional threshold and the first proportional threshold, then the k value remains unchanged; the k value after each adjustment is constrained between the minimum and maximum values.

[0026] Compared with the prior art, the beneficial effects of the present invention are:

[0027] 1. By continuously sampling high-frequency wind speed and analyzing fixed-window slices, we innovatively construct three core dynamic feature vectors: wind speed disorder, disturbance frequency band, and rate of change skewness. This enables refined, digital, and multi-dimensional quantitative identification of complex wind field conditions in ports. The disorder index of wind speed fluctuations accurately quantifies the degree of wind field disorder, intuitively determining the reliability of temperature sensor readings. The energy concentration frequency band of wind speed fluctuations accurately distinguishes the speed of wind field disturbances, providing a basis for controller acceleration tracking or active avoidance strategies. The distribution skewness direction of the rate of change of wind speed accurately identifies the asymmetric characteristics of gusts and predicts the trend of superheat rise and fall.

[0028] 2. Achieve dynamic adaptive construction of the superheat control benchmark for port heat pumps. By coupling wind field disorder with inherent equipment parameters, calculate the thermal reliability weight, accurately quantify the distortion of the measured data from the temperature sensing bulb, and innovatively introduce a wind speed-coupled first-order linear extrapolation algorithm for superheat. Utilize the wind speed advance change characteristics to compensate for the phase lag of the temperature sensing bulb's thermal inertia, obtaining a trend prediction value that closely matches the actual operating conditions. Through weighted fusion of measured and predicted values, achieve dynamic switching of the control benchmark: when the wind field is stable, prioritize the measured value to ensure control accuracy; when the wind field is disordered, switch to the trend prediction value to avoid lag distortion errors.

[0029] 3. Achieve full-condition adaptive modulation of EEV control action cycle and adjustment step size. By dynamically calculating the active sleep window through the wind field disturbance cycle, the controller avoids the resonance range of wind field disturbances, freezes ineffective adjustment actions, and completely solves the control resonance and system oscillation problems caused by fixed cycle, significantly improving the anti-interference capability of the system under complex wind fields. At the same time, asymmetric step size adjustment is achieved based on the skew direction of wind speed fluctuations. For impact gusts, the liquid supply is increased in advance, and for suction gusts, the liquid supply is reduced in advance. The asymmetric change trend of superheat is accurately predicted and adapted to, effectively suppressing superheat overshoot, eliminating liquid slugging failure, and protecting the core equipment of the compressor.

[0030] 4. Construct the only fully closed-loop adaptive iterative link in the entire system to realize the technical upgrade of control parameters from fixed preset to dynamic self-calibration. By accurately separating the wind speed disturbance interference and the thermal response component of EEV action, the matching degree between the current control cycle and the operating condition is accurately determined. The core coefficient k value of the dormant window is dynamically iteratively corrected, so that the EEV action cycle continuously tracks the port wind field rhythm and the dynamic phase difference of equipment thermal inertia. Attached Figure Description

[0031] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0032] Figure 1 This is a topological diagram of the present invention;

[0033] Figure 2This is a flowchart of the control logic of the present invention. Detailed Implementation

[0034] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0036] Please see Figures 1-2 As shown, a port green smart heating system based on a distributed air source heat pump includes a wind field rhythm sensing sub-center. The wind field rhythm sensing sub-center performs wind field rhythm sensing, that is, it performs environmental interference analysis based on the geographical location of the port, and uses the temporal disorder and frequency domain characteristics of wind speed disturbance to dynamically modulate the controller's own "effective sampling reference" and "action rhythm" so that the control rhythm of the EEV (electronic expansion valve) is synchronized with the "real thermal boundary change rhythm" rather than with the "lagging temperature sensor reading".

[0037] The feature recognition unit performs multi-dimensional dynamic feature vector acquisition.

[0038] Based on the wind speed sensor installed on the air inlet side of the heat pump unit, the instantaneous wind speed sequence of the outdoor unit's air inlet side is continuously collected at a sampling frequency of 10 times per second and recorded as v(t). At the same time, the thermal inertia time constant τ of the temperature sensing bulb stored locally in the heat pump unit is read. sensor (This parameter is determined by the material, wall thickness and filling medium of the temperature sensing bulb, and is an inherent property of the equipment. Its value range is 15 to 30 seconds, and its value range is the third range.)

[0039] The wind speed sequence v(t) is sliced ​​using a sliding window, with a window length L of 30 seconds (equal to twice the thermal inertia time constant of a typical temperature sensing bulb). Within each window, the following three features are extracted from the 300 wind speed sampling points in the window:

[0040] Feature 1: The disorder index of wind speed fluctuations, denoted as H vThis indicator is used to quantify the unpredictability and disorder of wind speed changes. The specific calculation method is as follows: the wind speed sequence in the window is arranged in chronological order to form a phase space, the embedding dimension is 2, the similarity tolerance (first preset value) is 0.2 times the standard deviation of wind speed in the window (second preset multiple), and the mean of the log-conditional probability of the new state generated under this tolerance is negative.

[0041] When the port is in a steady laminar flow state, H v When the value approaches a low value of 0.1 to 0.3, and gusts occur frequently with random fluctuations in wind speed, H... v The value can exceed 1.0. The physical meaning of this feature is that high disorder means that the current temperature sensor reading is superimposed with multiple random perturbations, and the reliability is extremely low. It is necessary to actively reduce the weight of this reading in the control decision.

[0042] Feature 2: The energy concentration frequency band of wind speed fluctuations, denoted as f p This index is used to perform a discrete Fourier transform on the wind speed sequence within a window, calculate the power spectral density, and extract the frequency value corresponding to the peak value of the power spectral density, characterizing the dominant rhythm of the wind speed disturbance—if f p If p falls within the 0.5–2 Hz range, it indicates that the disturbance is a high-frequency cluster gust. If f p A value below 0.05 Hz indicates that the disturbance is a slow drift, such as the alternation of sea and land winds. This feature is used to determine the relative relationship between the wind field change rhythm and the hysteresis time constant of the temperature sensing bulb, thereby deciding whether the controller should "accelerate tracking" or "actively avoid".

[0043] Feature 3: The skewness of the distribution of the rate of change of wind speed, denoted as S. dvdt This index is used to calculate the first derivative of wind speed with respect to time, dv / dt, at each moment within the window. Then, the ratio of the third central moment of this derivative sequence to the cube of the standard deviation is calculated as an index of skewness, characterizing the asymmetry between wind speed acceleration and deceleration processes. If S... dvdt A positive value indicates a "sudden increase and gradual decrease" in wind speed (impact gusts, with a steep leading edge and a gentle trailing edge). This type of gust will cause a sudden increase in evaporator heat exchange, and the actual superheat will soon rise rapidly. If S dvdt A negative value indicates that the wind speed "rises slowly and falls suddenly" (a suction-type gust, the heat exchange of the evaporator suddenly weakens, and the actual superheat is about to drop rapidly). If it approaches zero, it indicates that the wind speed fluctuation is symmetrical. This feature is used to predict the trend direction of superheat changes and provides a basis for the "asymmetric throttling" of EEV.

[0044] Effective benchmark building blocks, utilizing the obtained disorder index H v and energy concentration frequency band f pThe reliability of the current temperature sensor reading is dynamically assessed, and the use of "current measured value" or "trend extrapolation value" as the control basis is determined according to the level of reliability.

[0045] First, according to the disorder index H v and the inherent time constant τ of the temperature sensing bulb sensor The ratio of the thermodynamic reliability weight W to the window length L is used to calculate the thermodynamic reliability weight W of the current temperature sensor reading. trust The calculation relationship is as follows:

[0046] ;

[0047] Where λ is the preset scaling factor (range 0.5~1.5, initially set to 1.0, range is the first range), and L is the sliding window length in the feature recognition unit (fixed value 30 seconds, equal to the fourth preset value).

[0048] When wind speed is highly erratic (H) v When W is high trust An exponential decay indicates an increase in the deviation between the measured superheat of the current temperature sensing bulb and the actual thermal state at the evaporator outlet, thus reducing the reliability of the reading.

[0049] At the same time, read the measured superheat value SH of the evaporator outlet temperature sensing bulb. raw (t), and read the measured superheat value SH from the previous moment (1 second interval). raw (t-1), calculate the rate of change of the measured superheat: dSH raw / dt=SH raw (t)-SH raw (t-1).

[0050] Using this rate of change and the wind speed rate of change dv / dt, a first-order linear extrapolation is performed on the current superheat to obtain the extrapolated superheat trend value SH. pred (t):

[0051] ;

[0052] Where Δt is the sampling interval (1 second, the fifth preset value), and β is the preset coupling coefficient (the value range is 0.2~0.8, which represents the transfer gain of wind speed change on the evaporator heat transfer coefficient, the initial default value is 0.5, and the value range is the second range).

[0053] The physical meaning of this extrapolation is: to use the rate of change of wind speed to correct the trend of thermal state changes that the temperature sensor has not yet detected, and to compensate for the phase lag caused by thermal inertia.

[0054] According to W trust The size of the current measured superheat value S Hraw(t) and trend extrapolation value SH pred (t) is weighted and fused to obtain the effective superheat deviation ΔSH eff (t):

[0055] ;

[0056] Among them SH target Set the target superheat value at the evaporator outlet (automatically given by the system operating conditions, usually 5~10 Kelvin);

[0057] The essence of this weighted fusion is that when the wind field is chaotic, the controller actively switches the control basis from "lagging measured values" to "trend prediction values ​​based on wind speed correction" to avoid amplifying the lag error of the temperature sensor into control deviation.

[0058] The controller's action parameter self-modulation unit utilizes the obtained energy concentration frequency band f p Calculate the active sleep window T of EEV dwell This allows the EEV's decision-making rhythm to proactively avoid phase resonance with wind field disturbances and freeze unnecessary adjustment actions within the dormancy window.

[0059] By f p Calculate the energy concentration period T of wind field disturbance disturb =1 / f p .

[0060] Define the active sleep window T of the EEV dwell It is 0.6 to 0.8 times the period of the disturbance, i.e., T dwell =k×T disturb Where k ranges from 0.6 to 0.8 (the specific value is dynamically adjusted by subsequent closed-loop calibration of the unit, and the initial default value is 0.7).

[0061] The T dwell The physical meaning is: after the controller completes one EEV opening adjustment, it enters a period of length T. dwell The "observation dormancy period" is during which the controller does not respond to any changes in superheat deviation. It only samples the superheat value once at the end of the dormancy period to determine whether to execute the next action. The purpose is to maintain a fixed phase difference between the EEV's operating cycle and the wind speed disturbance period, preventing the EEV from being driven by the lag signal from the temperature sensor and striking the peaks or troughs of wind field fluctuations, thus avoiding resonance amplification. v When T is below 0.3 (stable wind field), dwell Remove the restriction and restore the controller's minimum response interval (sampling and decision once per second).

[0062] At the same time, using the obtained distribution skew direction S dvdtThe sign of the variable dynamically corrects the directional asymmetry step ratio of the EEV opening adjustment.

[0063] Specifically, the base step size is set to ΔStep. base (Calibrated under the system's rated operating conditions, typically 0.5% to 1.0% of the valve's maximum opening).

[0064] When the distribution is skewed in direction S dvdt When the value is greater than 0.2 (first directional threshold) (impact gust, the actual superheat is about to rise), the step multiplier for increasing the liquid supply direction (valve opening direction) is set to 1.6 times (first multiplier), and the step multiplier for decreasing the liquid supply direction (valve closing direction) is set to 0.4 times (second multiplier), so that the controller increases the liquid supply in advance before the actual superheat rises.

[0065] When the distribution is skewed in direction S dvdt When the value is less than -0.2 (second direction threshold) (suction-type gust, the actual superheat is about to drop), reverse the operation, set the step multiplier for the valve closing direction to 1.6 times, and the step multiplier for the valve opening direction to 0.4 times, and reduce the liquid supply in advance to prevent liquid hammer.

[0066] When the distribution is skewed in direction S dvdt When the value is between -0.2 and 0.2, the wind speed fluctuation is considered to be symmetrical, and the step size multiplier for both opening and closing the valve remains unchanged at 1.0. Thus, the effective opening adjustment step size of the electronic expansion valve is determined by the basic step size and the corresponding direction multiplier.

[0067] The closed-loop calibration dynamic adjustment unit uses the actual superheat response waveform after the EEV action to infer the current T. dwell Whether the settings are reasonable and dynamically adjust the k value to form a fully closed-loop adaptive link;

[0068] After each EEV completes an opening adjustment, the measured value of the evaporator outlet superheat SH is continuously monitored. raw (t) in subsequent T dwell The change curve over time contains both the "thermal response component caused by EEV action" and the "thermal disturbance component caused by wind speed disturbance," so the two need to be separated.

[0069] The separation method is as follows: within the same time window, based on the real-time calculated wind speed disorder H... v The theoretical change in superheat caused solely by wind speed disturbance is estimated using the wind speed change rate dv / dt (using a simplified linear model of heat exchange on the air side of the heat pump evaporator: the change in superheat is positively correlated with the wind speed change rate, and the proportionality coefficient is determined by the evaporator geometric parameters and the current operating conditions). This theoretical change is then subtracted from the measured change in superheat, and the remaining part is the pure response component caused by the EEV action.

[0070] From the pure response component curve, the moment when the superheat reaches an extreme value (peak or trough) is extracted and denoted as T. This T represents the time difference from the start of the EEV action to the occurrence of the superheat response extreme value, characterizing the system's true response speed under the current operating conditions.

[0071] Compare T with the currently set T dwell Comparison:

[0072] If T ≥ 1.3 (first proportional threshold) times T dwell (If the actual response is significantly slower than expected, indicating that the temperature sensing bulb lag is more severe than predicted or that sudden wind speed changes occur earlier), then the k value will be reduced by 0.05 (first adjustment step), so that T in the next control cycle... dwell Shorten and accelerate the intervention pace of EEV;

[0073] If T ≤ 0.7 (second proportional threshold) times T dwell (The actual response was significantly faster than expected, indicating that the change in superheat was mainly caused by the direct impact of wind speed on the outer shell of the temperature sensing bulb, rather than a change in the actual thermal state.) Therefore, the k value will be increased by 0.05, making T the value for the next control cycle... dwell Extending the timeline allows the controller to proactively "become less sensitive" in order to filter out false thermal shocks;

[0074] If T is between 0.7 and 1.3 times T dwell Between these points, it is considered that the current beat is well matched, and the k value remains unchanged.

[0075] Each adjusted k value is constrained to the range of 0.5 to 1.0 to ensure T dwell The value of k should be no less than 0.5 times the disturbance period (to avoid being too aggressive and causing resonance) and no more than 1.0 times the disturbance period (to avoid being too sluggish and causing mismatch). The adjusted k value should be immediately applied to T in the controller action parameter self-modulation unit. dwell The calculation completes the entire closed-loop adaptive adjustment from "wind field perception - action execution - effect feedback - parameter correction".

[0076] This closed-loop link enables the EEV's active sleep window T dwell By continuously tracking the dynamic phase difference between the current wind field disturbance beat and the system's thermal inertia, even if the temporal characteristics of the wind speed disturbance change with port tides or seasonal variations, the controller's action beat can be synchronously adjusted to always maintain the best matching relationship with the actual thermal boundary changes.

[0077] In summary, this system completely solves the deficiencies of lacking wind field feature recognition capabilities, ambiguous operating condition judgment, and lacking advance prediction capabilities. It provides accurate and quantitative operating condition data support for subsequent control benchmark correction and action parameter modulation, enabling precise matching of control strategies with real-time wind field conditions at the port. Furthermore, it addresses the problems of traditional control benchmarks being singular, lagging, and prone to distortion by eliminating control deviations caused by the thermal inertia of the sensing bulb at the source, significantly improving the accuracy and stability of superheat control under complex wind fields.

[0078] Meanwhile, under stable wind conditions, it automatically restores the high-frequency response mode, taking into account both the control accuracy under normal operating conditions and the operational stability under complex operating conditions. It comprehensively optimizes the EEV adjustment logic, effectively solving the problems of traditional control parameters being fixed, unable to adapt to seasonal and tidal changes, and poor long-term operational adaptability. It ensures that the system always maintains the optimal control matching state under all seasons, all tides, and all wind conditions, continuously optimizing the system's control accuracy, operational energy efficiency, and stability, and realizing the intelligent adaptive operation of the port heat pump heating system.

[0079] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A port green and intelligent heating system based on a distributed air source heat pump, comprising a wind farm rhythm sensing sub-center, characterized in that, The wind field rhythm sensing sub-center includes a feature recognition unit, an effective reference construction unit, a controller action parameter self-modulation unit, and a closed-loop calibration dynamic adjustment unit; The feature recognition unit is used to collect the instantaneous wind speed sequence of the air inlet surface of the heat pump unit at a preset sampling frequency, slice the sequence into sliding windows of a preset window length, and extract the following three features in each window: The disorder index of wind speed fluctuations, the frequency band of energy concentration of wind speed fluctuations, and the skew direction of the distribution of wind speed change rate. The effective benchmark construction unit is used to dynamically evaluate the reliability of the current temperature sensing bulb reading based on the disorder index and energy concentration frequency band, and to weight and fuse the current measured superheat value with the trend extrapolation value to obtain the effective superheat deviation. The controller action parameter self-modulation unit is used to calculate the wind field disturbance period based on the energy concentration frequency band, and calculate the active dormancy window of the electronic expansion valve based on the preset proportional coefficient k. At the same time, it dynamically corrects the direction step multiplier of the electronic expansion valve opening adjustment according to the sign of the distribution skew direction. The closed-loop calibration dynamic adjustment unit is used to monitor the superheat response curve after each action of the electronic expansion valve, extract the time when the superheat response extreme value occurs, dynamically correct the k value based on the comparison result between the time and the current active sleep window, and feed back the corrected k value to the controller action parameter self-modulation unit.

2. A port green and intelligent heating system based on a distributed air source heat pump according to claim 1, characterized in that, The wind speed sequence within the window is arranged in chronological order to form a phase space. The embedding dimension is set to a first preset value, and the similarity tolerance is set to a second preset multiple of the standard deviation of the wind speed within the window. The mean of the log-conditional probability of the new state generated under this tolerance is calculated and taken as a negative value.

3. A port green and intelligent heating system based on a distributed air source heat pump according to claim 1, characterized in that, The method for extracting the concentrated energy frequency band is as follows: perform a discrete Fourier transform on the wind speed sequence within the window, calculate the power spectral density, and extract the frequency value corresponding to the peak value of the power spectral density.

4. A port green and intelligent heating system based on a distributed air source heat pump according to claim 1, characterized in that, Calculate the first derivative of wind speed with respect to time at each moment within the window, and then calculate the ratio of the third central moment of the first derivative sequence to the cube of the standard deviation. A positive value of this ratio indicates a sudden rise and fall of wind speed in an impact gust, a negative value indicates a slow rise and fall of wind speed in a suction gust, and a value approaching zero indicates symmetrical wind speed fluctuations.

5. A port green and intelligent heating system based on a distributed air source heat pump according to claim 1, characterized in that, In the effective benchmark construction unit, the thermal reliability weight W of the current temperature sensor reading is... trust According to the disorder index H v The thermal inertia time constant τ of the temperature sensing bulb sensor The ratio of the preset window length L to the given value is calculated as follows: , where λ is the preset scaling coefficient; The trend extrapolation value SH pred (t) Calculated based on the current measured superheat value, the rate of change of the measured superheat value, and the rate of change of wind speed, using the following formula: Where Δt is the sampling interval and β is the preset coupling coefficient; The effective superheat deviation is calculated using the following formula: , of which SH target Set the target value for superheat.

6. A port green and intelligent heating system based on a distributed air source heat pump according to claim 1, characterized in that, In the controller action parameter self-modulation unit, the active sleep window T dwell =k×T disturb T disturb =1 / f p f p This refers to the energy concentration frequency band; When the distribution skew direction is greater than the first direction threshold, the valve opening direction step size multiplier is set as the first multiplier and the valve closing direction step size multiplier is set as the second multiplier. When the distribution skew direction is less than the second direction threshold, the valve closing direction step size multiplier is set as the first multiplier and the valve opening direction step size multiplier is set as the second multiplier. When the distribution skew direction is between the second direction threshold and the first direction threshold, the valve opening and closing step multipliers are both kept at 1, and the effective opening adjustment step of the electronic expansion valve is determined by the basic step and the corresponding direction multiplier.

7. A port green and intelligent heating system based on a distributed air source heat pump according to claim 1, characterized in that, In the closed-loop calibration dynamic adjustment unit, the time T at which the extreme value of the response occurs is compared with the current T. dwell Comparison: If T ≥ first proportional threshold × T dwell Then, the value of k is reduced by the first adjustment step; if T ≤ the second proportional threshold × T dwell If T is between the second proportional threshold and the first proportional threshold, then the k value will be increased by the first adjustment step size; if T is between the second proportional threshold and the first proportional threshold, then the k value will remain unchanged. Each adjusted k value is constrained between the minimum and maximum values.

Images