Pulse heating aircraft deicing method and system

Abstract

The invention relates to the technical field of aircraft electric heating deicing, in particular to a pulse heating aircraft deicing method and system.The method comprises the following steps that the ice layer thickness, the ice layer type, the airspeed and the attack angle of each aircraft surface area are collected, the inherent pneumatic sensitivity coefficient of each area is called, the aerodynamic penalty index of each area is established, and the aerodynamic penalty index of each area is calculated; and according to the values of the aerodynamic penalty indexes of the areas, performing descending order arrangement to generate a deicing area priority sequence. Through multi-parameter joint evaluation of the ice layer thickness, the ice layer type, the airspeed and the incidence angle, the heating resource distribution sequence is optimized, energy waste is avoided, in the local heating execution stage, temperature change slope real-time monitoring is adopted to replace fixed time control, the actual thermal response behavior of the ice layer can be dynamically responded, and the real-time thermal response performance of the ice layer is improved. The initial phase change process of the ice crystal is identified through a phase change starting slope threshold value, and then a medium-power pulse control mode is switched, so that a steady-state regulation heating path is formed, and the phase change control continuity is strengthened.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aircraft electric heating deicing, in particular to a pulse heating aircraft deicing method and system. BACKGROUND

[0002] Aircraft icing can destroy the original flow field, cause lift to decrease and resistance to increase, damage the stability and safety of flight control, and bring major safety hazards to flight safety. Aircraft icing has become a focus of the current aviation industry. In order to ensure flight safety, various deicing methods have been developed. Among them, electric heating deicing technology has become an important deicing method due to its high operability and no damage to flight aerodynamic performance. However, the existing electric heating deicing technology mainly adopts fixed power or timing cycle strategy, ignores the differences in aerodynamic characteristics of icing areas, changes in ice layer structure, and thermal feedback dynamics in the deicing process, resulting in a lack of basis for regional processing sequence, and prone to priority configuration errors, thereby causing a reaction lag on key aerodynamic surfaces. At the same time, the existing technology mainly uses preset time or fixed program control, which is easy to cause incomplete deicing of some areas or energy redundancy, and the heating opportunity is difficult to match the actual ice layer thermal response rate. Therefore, improvement is needed. SUMMARY

[0003] The purpose of the present application is to solve the problems existing in the prior art and to provide a pulse heating aircraft deicing method and system.

[0004] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme, a pulse heating aircraft deicing method, comprising the following steps: S1, collecting the ice layer thickness, ice layer type, airspeed and angle of attack of each aircraft surface area, retrieving the inherent aerodynamic sensitivity coefficient of each area, establishing an aerodynamic penalty index of each area, arranging the values of the aerodynamic penalty index of each area in descending order, and generating a deicing area priority sequence; S2, according to the deicing area priority sequence, selecting the heating element of the first area in the sequence to apply an initial heating pulse, and simultaneously calculating the real-time temperature change slope, continuously determining the real-time temperature change slope, and when it falls to a preset phase change starting slope threshold, establishing an ice crystal phase change starting feature point; S3, according to the ice crystal phase change starting feature point, adjusting the pulse parameters applied to the heating element to form a phase change maintenance pulse sequence, and continuously calculating the temperature change slope of the temperature time data during the application of the phase change maintenance pulse sequence, updating the real-time temperature change slope, and when the real-time temperature change slope jumps from a state close to zero and exceeds a melting end slope threshold, establishing a regional melting completion identifier; S4, according to the region melting completion identifier, stop supplying pulse energy to the heating element of the current region, remove the processed region in the deicing region priority sequence, obtain an updated deicing region priority sequence, select the region with the highest ranking in the updated deicing region priority sequence as the next target, and cyclically execute the steps S2 and S2 to establish the next cycle heating target instruction.

[0005] Preferably, the step of obtaining the deicing region priority sequence is: analyze each aircraft surface region identifier and collect the ice layer thickness value, ice layer type identifier, airspeed value and angle of attack value one by one, call the aerodynamic sensitivity coefficient value matched with the region number, and introduce the region characteristic length value and stall angle of attack value, normalize to generate a region parameter list; According to the region parameter list, calculate the region aerodynamic penalty index; According to the region aerodynamic penalty index, arrange the region aerodynamic penalty index in descending order from high to low, if the absolute value of the difference between the region aerodynamic penalty indexes of adjacent regions is less than a preset criterion, then arrange according to the absolute value of the angle of attack, if still the same, arrange according to the airspeed, form a deicing region priority sequence.

[0006] Preferably, the step of obtaining the real-time temperature change slope is: According to the deicing region priority sequence, locate the first region and apply an initial heating pulse to the region heating element, continuously read the time stamp and temperature value and eliminate abnormal fluctuation points, complete data alignment according to time stamp order to obtain a temperature time data sequence; According to the temperature time data sequence, calculate the real-time temperature change slope.

[0007] Preferably, the step of obtaining the ice crystal phase transition starting feature point is: According to the real-time temperature change slope, compare it with the phase transition starting slope threshold value point by point, when the real-time temperature change slope is first detected to be less than or equal to the phase transition starting slope threshold value, read the corresponding temperature value and time stamp and combine them into a pair of fixed values to generate an ice crystal phase transition starting feature point.

[0008] Preferably, the step of obtaining the phase transition maintenance pulse sequence is: According to the ice crystal phase transition starting feature point, locate the phase transition maintenance switching time, terminate the short-time pulse and load the medium power parameter, increase the duty cycle from the initial duty cycle to the target duty cycle, record the effective time stamp and parameter group, and generate a phase transition maintenance pulse sequence.

[0009] Preferably, the step of obtaining the region melting completion identifier is: According to the phase change maintaining pulse sequence, the temperature time data is sequentially processed in a fixed length time window during pulse application, and a real-time temperature change slope is updated; According to the real-time temperature change slope, a first jump point after a monotonic discrimination and tracking of a near-zero interval is continuously determined with a melting end slope threshold, and a region melting completion identifier is formed when the real-time temperature change slope is greater than the melting end slope threshold.

[0010] Preferably, the obtaining step of the updated deicing region priority sequence is: According to the region melting completion identifier, a stop supply instruction is issued to the heating element of the current region, the pulse energy output is turned off, the supply current value and the duty cycle value are read and confirmed to be zero, the stop supply effective timestamp, the last pulse parameter and the current region identifier are recorded, and a stop supply confirmation state is generated; According to the stop supply confirmation state, the current region identifier is searched and deleted in the deicing region priority sequence, the remaining regions are rearranged in descending order of the region aerodynamic penalty index, the rearrangement timestamp and the sequence checksum are recorded, and the updated deicing region priority sequence is generated.

[0011] Preferably, the obtaining step of the next cycle heating target instruction is: According to the updated deicing region priority sequence, the highest ranked region identifier is selected and filled with the initial heating pulse parameter, the temperature sampling period and the phase change start slope threshold, and the steps S2 and S2 are executed in a loop to form the next cycle heating target instruction.

[0012] The application also provides an aircraft deicing system, comprising: A perception evaluation module is configured to collect the ice layer thickness, the ice layer type, the airspeed and the angle of attack of each aircraft surface region, retrieve the inherent aerodynamic sensitivity coefficient of each region, establish the aerodynamic penalty index of each region, arrange the regions in descending order according to the values of the aerodynamic penalty index, and generate a deicing region priority sequence; A heating start module is configured to select the heating element of the first region in the sequence according to the deicing region priority sequence, apply an initial heating pulse to the heating element, continuously determine the real-time temperature change slope, and establish an ice crystal phase change start feature point when the real-time temperature change slope falls to a preset phase change start slope threshold. A phase change maintaining module is configured to adjust the pulse parameters applied to the heating element according to the ice crystal phase change start feature point, form a phase change maintaining pulse sequence, continuously perform temperature change slope operation on the temperature time data during application of the phase change maintaining pulse sequence, update the real-time temperature change slope, and establish a region melting completion identifier when the real-time temperature change slope jumps from a near-zero state and exceeds a melting end slope threshold. A cycle control module is configured to stop supplying pulse energy to the heating element of the current region according to the region melting completion identifier, remove the processed region from the deicing region priority sequence, obtain an updated deicing region priority sequence, select the region with the highest priority in the updated deicing region priority sequence as the next target, and establish a next cycle heating target instruction.

[0013] Compared with the prior art, the application has the advantages and positive effects that: The application optimizes the heating resource allocation sequence by jointly evaluating the ice layer thickness, ice layer type, air speed and angle of attack, avoids energy waste, and uses the temperature change slope to monitor in real time during the local heating execution phase instead of fixed time control, which can dynamically respond to the actual thermal response behavior of the ice layer, identify the initial phase change process of ice crystals through the phase change start slope threshold, and then switch to the pulse control mode of medium power to form a steady heating path and strengthen the continuity of phase change control. The temperature change trend during the phase change process is continuously sampled and the melting completion node is identified based on the slope jump, the deicing region priority sequence is updated in a timely manner based on the region melting completion identifier, the processing progress is synchronized in real time, the real-time performance of the deicing response is improved, and the increase in power consumption caused by lagging or redundant heating strategies is avoided. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 The application provides a step schematic diagram. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical scheme and advantages of the application clearer, the application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the application and do not limit the application.

[0016] Please refer to Figure 1 The application provides a technical scheme, a pulse heating aircraft deicing method, comprising the following steps: S1, collecting the ice layer thickness, ice layer type, air speed and angle of attack of each aircraft surface region, retrieving the inherent aerodynamic sensitivity coefficient of each region, establishing the aerodynamic penalty index of each region, arranging the values of the aerodynamic penalty index of each region in descending order, and generating a deicing region priority sequence; S2, according to the deicing region priority sequence, selecting the heating element of the first region in the sequence to apply an initial heating pulse, and calculating the real-time temperature change slope, continuously determining the real-time temperature change slope, and when it falls to a preset phase change start slope threshold, establishing an ice crystal phase change start feature point; S3, according to the ice crystal phase transition starting feature point, adjusting the pulse parameters applied to the heating element, forming a phase transition maintenance pulse sequence, during the application of the phase transition maintenance pulse sequence, the temperature change slope operation is continuously performed on the temperature time data, the real-time temperature change slope is updated, when the real-time temperature change slope jumps from the state approaching to zero and exceeds the melting end slope threshold, the regional melting completion identifier is established; S4, according to the regional melting completion identifier, stopping supplying pulse energy to the heating element of the current region, and removing the processed region in the deicing region priority sequence to obtain an updated deicing region priority sequence, selecting the region with the highest ranking in the updated deicing region priority sequence as the next target, and cyclically executing S2 and S2 steps to establish the next cycle heating target instruction.

[0017] The obtaining step of the deicing region priority sequence is: The ice layer thickness value, ice layer type identifier, airspeed value and angle of attack value of each aircraft surface region identifier are analyzed and collected one by one, the aerodynamic sensitivity coefficient value matched with the region number is called, and the region characteristic length value and stall angle of attack value are introduced, normalized to generate a region parameter list; According to the region parameter list, the region aerodynamic penalty index is calculated, and the calculation formula is: ; Wherein, the region aerodynamic penalty index, the region aerodynamic sensitivity coefficient value, the region ice layer type influence factor, the region ice layer thickness value, the region characteristic length value, the ice layer thickness effect index, the air density value, the region airspeed value, the reference dynamic pressure value, the absolute value of the region angle of attack, the region stall angle of attack value, , , the weight coefficient determined by experiment; According to the region aerodynamic penalty index, the region aerodynamic penalty index is arranged in descending order from high to low, if the absolute value of the difference between the region aerodynamic penalty indexes of adjacent regions is less than a preset criterion, the angle of attack absolute value is sorted, if it is still the same, the airspeed is sorted, forming the deicing region priority sequence.

[0018] Specifically, according to the analysis of each aircraft surface area identification, such as wing leading edge A1 area, wing leading edge A2 area, horizontal tail B1 area, vertical tail C1 area, etc., start the multi-sensor data acquisition process, for each identified area, the ultrasonic sensor array deployed at a frequency of 50 Hz for pulse echo detection, by calculating the time difference of ultrasonic reflection at the ice layer and the skin interface, combined with the sound speed in ice (about 3800 meters per second) to convert the ice layer thickness value, the synchronous deployment of optical ice layer detector by analyzing the reflection spectrum characteristics, the ice layer type is identified as clear ice, frost ice or mixed ice, and this classification information is recorded, at the same time, the airspeed and angle of attack values are read in real time from the ARINC 429 data bus of the aircraft, these two parameters directly reflect the current flight state of the aircraft, then, access the aerodynamic characteristics database stored in the non-volatile memory of the onboard computer, according to the current processing area identification (such as A1 area) as the index, call the aerodynamic sensitivity coefficient value, regional characteristic length value (such as wing chord length) and regional stall angle of attack value of the region determined by computational fluid dynamics (CFD) simulation and wind tunnel test in advance, these inherent parameters represent the sensitivity of each region to the aerodynamic performance degradation of icing, then, all the collected and retrieved dynamic and static values are normalized, using the minimum-maximum normalization method, linearly map each parameter value to the interval, the normalization formula is wherein is the original parameter value, and is the historical minimum and maximum value of the parameter in the entire flight envelope of the aircraft, for example, the ice layer thickness is set to 0 millimeter, is set to 50 millimeters, the airspeed is 80 meters per second, is 250 meters per second, finally, each region identification and its corresponding all original values and normalized values are integrated into a structured data record, and all region data records together constitute the region parameter list.

[0019] In the formula for calculating the regional aerodynamic penalty index, the physical properties of icing (thickness, type), the geometric and aerodynamic characteristics of the region (sensitivity, characteristic length, stall angle of attack), and the current flight state (airspeed, angle of attack) are integrated into a unified evaluation framework, which not only considers the direct impact of ice layer thickness , but also simulates the nonlinear disturbance effect of ice shape on airflow through the ratio of characteristic length and the index , and introduces the dynamic pressure term and the angle of attack term The coupling of the two factors makes the penalty index able to dynamically reflect the feature that the risk of the aircraft increases sharply when approaching the stall boundary, thus achieving a more intelligent and risk-oriented allocation of deicing resources.

[0020] is the aerodynamic sensitivity coefficient of the region, which represents the influence of icing on the specific region of the aircraft on the overall aerodynamic performance (such as lift loss, drag increase). It is obtained based on computational fluid dynamics (CFD) simulation and high-fidelity wind tunnel test in the design phase of the aircraft. In the simulation model, a standard-shaped simulated ice type is applied to each key region of the aircraft (such as wing root, wing tip, horizontal tail, etc.), and then the weighted sum of the resulting decrease in total lift coefficient and increase in total drag coefficient is calculated. The performance degradation of this region is compared with the most severe performance degradation among all regions, and the value obtained after normalization is , which is between 0 and 1. The higher the value, the more sensitive the region is. For example, the center region of the wing leading edge is most sensitive to lift, so its value is usually close to 1, while the value of the region such as the fuselage belly is lower. A specific example is as follows: CFD simulation is performed on a wing tip region, and it is found that icing of this region leads to a 5% decrease in total lift coefficient and an 8% increase in total drag coefficient. The most sensitive region (middle part of the wing leading edge) icing leads to an 8% decrease in lift and a 10% increase in drag. If the weights of lift and drag are 0.5 each, the comprehensive influence degree of the wing tip region is , and the influence degree of the most sensitive region is , then the value of the wing tip region is In this example, the of a key region of the wing leading edge is .

[0021] is the regional ice layer type influence factor, which is a dimensionless coefficient that quantifies the difference in the influence of different ice layer types on aerodynamic performance. Since non-numerical ice layer type identifiers (such as clear ice, frost ice) cannot be directly substituted into the formula for calculation, a quantitative mapping standard needs to be established. This standard is based on icing wind tunnel test data. Experiments show that smooth and dense clear ice causes much greater disturbance to airflow than rough and low-density frost ice. Therefore, by analyzing the influence curves of different ice types on the lift-drag ratio characteristics of airfoils in a large amount of experimental data, a baseline (usually frost ice) is set, and its value is set to 1.0. Then, according to the degree of aerodynamic penalty caused by other ice types, the relative coefficient is determined. For example, experiments show that clear ice causes a 1.4 times lift loss compared to frost ice, so its The value can be set as 1.4, and an example of acquisition is: through the on-board optical sensor, the current regional ice layer is identified as "clear ice", and the preset mapping table is consulted to obtain The value is 1.4.

[0022] is the regional ice layer thickness value, in meters (m), which is obtained by real-time measurement through ultrasonic or optical sensors installed on the surface of the aircraft skin. The sensor works continuously at a fixed frequency (for example, 10 Hz) and transmits measurement data to the de-icing control computer. The computer performs sliding average filtering on the continuous readings to eliminate single-pulse measurement noise and obtain a stable and representative thickness value. For example, in a certain sampling period, the sensor continuously measures thicknesses of 2.8 mm, 3.1 mm, 2.9 mm, 3.2 mm, and 3.0 mm. After 5-point sliding average, the obtained is mm, i.e. m.

[0023] is the regional characteristic length value, in meters (m), which is a fixed parameter related to the geometric shape of the aircraft and represents the macroscopic size of a specific region. For wing or tail surfaces, it is usually taken as the local chord length. For other components such as the nose, it is taken as the radius of curvature or a representative length. This parameter is determined during the aircraft design phase and is pre-stored in the on-board database and can be called through the region identifier. For example, for the main wing root region of a medium-sized passenger aircraft, the characteristic length may be m.

[0024] is the ice layer thickness effect index, which is a dimensionless empirical constant used to describe the non-linear effect of ice layer thickness growth on aerodynamic penalty. Generally, the impact of thin ice layer is relatively small, but as the thickness increases, its destructive effect on airflow will increase disproportionately sharply, The value of is usually greater than 1, and its specific value is obtained by power function fitting of wind tunnel test data at different thicknesses. For example, for the lift loss data points of a certain airfoil at different ice thicknesses, if it is found that the lift loss is proportional to the 1.2 power of , then is .

[0025] is the air density value, in kilograms per cubic meter (kg / m 3), which is calculated by the Air Data Computer (ADC) from the static pressure and total temperature at the current flight altitude, conforming to the International Standard Atmosphere (ISA), for example, at a cruising altitude of 8000 meters, the air density under standard atmospheric conditions is approximately kg / m 3 .

[0026] is the regional airspeed value, in meters per second (m / s), which is also provided by the Air Data Computer, representing the speed of the aircraft relative to the surrounding air, using the True Airspeed (TAS) provided by the ADC as a unified , for example, if the current True Airspeed of the aircraft is 220 m / s, then m / s.

[0027] is the reference dynamic pressure value, in Pascals (Pa), which is a reference value used to normalize the current dynamic pressure, so that the penalty index is comparable under different flight conditions, usually the dynamic pressure of the aircraft flying at its maximum cruising speed (for example kg / m 3 ) at standard sea level altitude (approximately m / s) is chosen as the reference value, the calculation formula is: , the setting process is: Pa.

[0028] is the regional absolute angle of attack value, in degrees (°), provided by the flight control system or the Air Data Computer, representing the angle between the chord line of the wing and the direction of the far-forward airflow, for example, the current angle of attack of the aircraft is .

[0029] is the regional stall angle of attack value, in degrees (°), which is the angle of attack at which a specific airfoil region can achieve the maximum lift coefficient in an ice-free state, it is an inherent aerodynamic characteristic parameter, pre-stored in the onboard database, exceeding this angle of attack, the lift will drop sharply, leading to stall, for example, for the selected airfoil region, the stall angle of attack in a clean state is . .

[0030] , , are experimentally determined weight coefficients, these dimensionless coefficients are used to adjust the relative importance of different terms in the formula, adjust the penalty weight of ice layer geometry (thickness), adjust the penalty weight of the current aerodynamic load, and Then the risk amplification effect of the aircraft approaching the stall state is specially adjusted, and their settings are not subjective settings, but are obtained through an optimization process, the specific steps are: collect a large amount of flight data or high-fidelity simulation data, which contains various icing conditions and their corresponding actual flight risk levels (rated by test pilots or quantified by performance degradation), and then calculate the The difference between the ranking and the actual risk ranking is the objective function, and the genetic algorithm or particle swarm optimization algorithm is used for iterative optimization, and finally a set of optimal weight coefficients is determined, and the weight value determined through this process is: , , .

[0031] According to the above parameters, the calculation is as follows: Substitute the obtained numerical value into the formula: .

[0032] .

[0033] m.

[0034] m.

[0035] .

[0036] kg / m 3 .

[0037] m / s.

[0038] Pa.

[0039] ; ; ; ; ; Calculation process: ; ; ; ; ; ; ; ; ; ; ; ; The result shows that the current aerodynamic penalty index of the key area of the wing leading edge is 0.21536, and after obtaining the values of all icing areas, the value will be used for sorting, the higher the value of the area, the greater the threat of the icing state to flight safety, and it should be given higher deicing priority, for example, if the calculated value of another area (such as the horizontal tail) is 0.35, it means that the deicing demand of the horizontal tail is more urgent than that of the wing area, and the horizontal tail will be processed first.

[0040] According to the calculated aerodynamic penalty index of each area, all areas and their corresponding penalty index values are taken as a set, for example {A1 area: 0.215, A2 area: 0.211, B1 area: 0.350, C1 area: 0.180}, first execute the first level sorting rule, that is, strictly according to the value of the penalty index in descending order, get the preliminary sorting {B1 area: 0.350, A1 area: 0.215, A2 area: 0.211, C1 area: 0.180}, then, the second level sorting rule is started, that is, the penalty index of adjacent areas is compared, and it is checked whether the absolute value of their difference is less than the preset criterion, which is set based on the measurement accuracy of the system sensor and the calculation uncertainty of the model, aiming to avoid unnecessary priority frequent switching caused by small value fluctuations, the setting process is: statistics the fluctuation standard deviation of the penalty index value of each area in the stable flight stage in the last 100 flights, take 1.5 times of the average value as the criterion, for example, the average standard deviation is calculated to be 0.008, and the criterion is set to In the above sequence, the program checks the difference between A1 area and A2 area, ​​Since 0.004 is less than the criterion 0.012, the secondary sorting logic is triggered, at this time, the program calls the absolute value of the angle of attack of each A1 and A2 area, for example, the angle of attack of A1 area is 4.5 degrees, and the angle of attack of A2 area is 4.8 degrees, because the larger the angle of attack usually means that the aerodynamic state is closer to the critical state, so the priority of A2 area with larger angle of attack should be promoted, therefore the sequence is adjusted to {B1 area: 0.350, A2 area: 0.211, A1 area: 0.215, C1 area: 0.180}, if the absolute values of the angles of attack of the two areas are also the same, the third level sorting rule will be entered, that is, comparing the airspeed of the two areas, the area with higher airspeed is given higher priority because the icing growth rate may be faster or the energy demand is larger, after all the comparison and adjustment is completed, a stable and multi-factor considered deicing area priority sequence is finally obtained.

[0041] The acquisition step of the real-time temperature change slope is: According to the deicing area priority sequence, the first area is located and an initial heating pulse is applied to the area heating element, the time stamp and temperature value are continuously read and the abnormal fluctuation points are removed, the data alignment is completed in time stamp order, and the temperature time data sequence is obtained; According to the temperature time data sequence, the real-time temperature change slope is calculated, and the calculation formula is: ; Wherein, is the real-time temperature change slope at the kth time point, is the size of the sliding window, is the current calculation index, is the ith time stamp, is the ith temperature value.

[0042] Specifically, according to the previously obtained sequence of priorities of the deicing regions, the region with the highest ranking in the sequence is first located as the current processing target, for example, the B1 region (horizontal tail fin), and then an instruction is issued to the power controller of the heating element array integrated to the region to apply a preset initial heating pulse, the pulse being characterized by high power and short time, and the specific parameter settings being a pulse width modulation (PWM) frequency of 1 kHz, a duty cycle of 95%, and a duration of 3 seconds. At the moment of pulse application, a high-frequency data acquisition program is started synchronously to continuously read the output voltage of an embedded temperature sensor (for example, a Pt100 resistance temperature detector) adjacent to the heating element at a sampling rate of 100 Hz, and the internal clock of the onboard computer is called to obtain the corresponding timestamp. The voltage value is converted into a temperature value through a pre-stored calibration curve. Then, for each new temperature data point collected, an abnormal fluctuation elimination logic is executed. The logic first determines whether the absolute value of the difference between the new data point and the previous valid data point exceeds a threshold value calculated based on the maximum physical heating rate. For example, according to the heating element power and material heat capacity, the maximum temperature rise rate is calculated to be 40 degrees Celsius per second. Therefore, within a sampling interval of 0.01 seconds, the temperature change should not exceed 0.4 degrees Celsius. Any data point exceeding this change rate is marked as electrical noise interference and discarded. The retained data points are further smoothed by a three-point median filter to further eliminate small random fluctuations. Finally, the processed valid timestamp and temperature value are paired and stored in a first-in, first-out (FIFO) buffer in strict chronological order to form a continuous and clean temperature time data sequence.

[0043] In the real-time temperature change slope calculation formula, a limited data window is applied on the time axis. The design idea is to obtain the best linear trend line by least squares fitting of the temperature data within a small time period. The slope of the trend line is the most reliable estimate of the temperature change rate at the current time. Compared with the simple two-point difference method (i.e. ), the information of data points in the window is utilized to smooth the instantaneous measurement errors caused by sensor noise or small environmental fluctuations, thereby revealing the true physical trend of temperature change more clearly, especially in identifying the critical node of ice layer melting, where the energy shifts from increasing the temperature of the substance (sensible heat) to changing the phase (latent heat), resulting in a small but clear inflection point in the temperature rise rate.

[0044] The size of the sliding window is a dimensionless positive integer that determines the number of data points used to calculate the slope. Its selection is a key tuning process that needs to balance response speed and signal smoothness. A too small value will result in a slope that is sensitive to noise and fluctuates dramatically, while a too large This value introduces a delay, potentially missing the precise moment when the phase transition begins. The value was determined by conducting numerous de-icing heating experiments in a ground-based icing laboratory, recording temperature curves under different heating powers and ice conditions, and then using different... Values ​​(e.g., 5, 10, 15, 20) are used to replay historical data for calculations. The calculated slope curve is then compared with the actual melting initiation point observed by a high-speed camera, and the value that most quickly and stably reflects the slope decrease at the melting initiation point is selected. Values, for example, through experimental analysis, show that at a sampling rate of 100 Hz, the selected... (That is, analyzing data from the past 0.15 seconds) can control the detection delay to within 0.2 seconds at a noise level of 5%, therefore it is determined that... The value is 15.

[0045] The current calculation index is a dimensionless integer representing the position of the currently processed data point within the entire temperature time series. The calculation starts from... start.

[0046] This is the i-th timestamp, in seconds (s). This value comes from the temperature time data sequence generated in the previous step and represents the precise time when the i-th temperature data point was collected. It is provided by the high-precision clock of the onboard computer.

[0047] The i-th temperature value is in degrees Celsius (°C), which is also derived from the temperature time data series and is the effective temperature reading after outlier removal and smoothing.

[0048] Calculations based on parameters: Use the temperature-time data sequence obtained in the previous step, and set the sliding window size. Calculate the first real-time temperature change slope at each time point At this point, an index is needed. arrive The following is a sample of 15 data points: Timestamp (Unit: s): {0.06, 0.07, 0.08, 0.09, 0.10, 0.11, 0.12, 0.13, 0.14, 0.15, 0.16, 0.17, 0.18, 0.19, 0.20}.

[0049] temperature (Units °C): {-8.2, -7.8, -7.3, -6.9, -6.5, -6.0, -5.6, -5.2, -4.8, -4.3, -3.9, -3.5, -3.0, -2.6, -2.2}.

[0050] The sums required for the calculation are: ; ; ; ; ; Substitute these values into the formula: Molecule calculation: ; ; Denominator calculation: ; ; Slope Calculation: ; This result shows that at the time point 0.20 seconds, by linear regression analysis on the data of the past 0.15 seconds, the temperature rise rate is about 36.26 degrees Celsius per second, which reflects that in the initial heating stage, the ice layer and the aircraft skin are rapidly absorbing heat, and the temperature is rising steadily. The slope value will be continuously output for real-time comparison with the phase change starting slope threshold value. A high and stable positive slope value (such as 36.26) confirms that it is in the effective sensible heat heating stage and has not yet entered the phase change process. When the subsequent calculated slope value starts to continuously decrease and approaches a preset low threshold value, it indicates that the melting process is about to begin.

[0051] The steps for obtaining the ice crystal phase change starting feature point are: According to the real-time temperature change slope, compare it with the phase change starting slope threshold value point by point. When the real-time temperature change slope is first detected to be less than or equal to the phase change starting slope threshold value, read the corresponding temperature value and time stamp and combine them into a pair of fixed values to generate the ice crystal phase change starting feature point.

[0052] Specifically, based on the real-time calculated and continuously output data stream of the temperature change slope, a comparison program is initiated point-by-point with a preset phase change initiation slope threshold. This threshold is set based on experimental data supporting the physical process of ice-water phase change. Ideally, when ice formed from pure water reaches 0 degrees Celsius, the continuously input heat will be entirely used to break the molecular bonds of the ice lattice, at which point the temperature will no longer rise, and the temperature change slope should theoretically be zero. However, in actual aircraft de-icing scenarios, due to impurities in the ice layer, incomplete and uneven heating, and heat exchange with the surrounding cold air, the temperature will still rise slightly during melting, and the slope will not be absolutely zero. Therefore, the process for setting the phase change initiation slope threshold is as follows: In a ground laboratory simulation... In a simulated high-altitude, low-temperature environment, hundreds of heating and melting experiments were conducted on standard ice samples of different types (clear ice, frost ice) and thicknesses. High-precision thermocouples and data loggers were used to monitor temperature changes. The real-time temperature change slope values ​​observed visually within 0.5 seconds before and after the first drop of melt water appeared in each experiment were recorded, forming a dataset containing hundreds of slope values. Statistical analysis was performed on this dataset, removing the top 5% and bottom values ​​to eliminate interference from abnormal operating conditions. Then, the mean and standard deviation of the remaining data were calculated. Finally, the phase change initiation slope threshold was set to "the mean plus one standard deviation." For example, if the calculated mean is 0.4 degrees Celsius per second and the standard deviation is 0.15 degrees Celsius per second, the threshold is set to... In real-time comparisons, the program maintains a Boolean state flag, "Phase transition detection triggered," initially false, for each newly calculated real-time temperature change slope, such as... All execute the judgment "if" "If the value is less than or equal to 0.55 and 'phase change detection trigger' is false", once this condition is met for the first time, the program immediately sets the "phase change detection trigger" flag to true and simultaneously reads the timestamp and temperature value corresponding to that point from the current temperature-time data sequence. For example, when the calculated value is... When the conditions are met, the program reads the timestamp as 1.50 seconds and the temperature as -0.05 degrees Celsius, and then combines these two values ​​into a fixed binary pair (1.50, -0.05) to generate the ice crystal phase transition initiation feature point.

[0053] The steps for obtaining the phase transition sustaining pulse sequence are as follows: Based on the initial characteristic point of ice crystal phase transition, locate the phase transition sustaining switching time, terminate the short-time pulse and load medium power parameters, adjust the duty cycle from the initial duty cycle to the target duty cycle, record the effective timestamp and parameter group, and generate the phase transition sustaining pulse sequence.

[0054] Specifically, according to the ice crystal phase transition starting feature point obtained in the previous step, the time information contained therein, such as 1.50 seconds, is directly used as the phase transition maintenance switching time, and the control system immediately issues an instruction to the power controller of the current area heating element, terminates the previously applied high-power short pulse, and specifically sets the duty cycle of the pulse width modulation (PWM) signal to 0% instantaneously, cutting off the energy supply, and then loads a set of preset medium-power parameters, the core of which is the target duty cycle, the value of which is not fixed but dynamically determined through a three-dimensional lookup table pre-stored in the on-board computer memory, taking the ice layer thickness of the current area, the real airspeed of the external air, and the external air temperature as input indexes, and outputting a duty cycle value corresponding to the optimal maintenance power. The data source of this lookup table comes from a large number of ground icing wind tunnel experiments and thermodynamic simulation calculations, aiming to match a power level that can exactly compensate for the latent heat required for ice layer melting and heat loss to the environment for different icing and flight conditions. For example, for a condition of ice layer thickness of 3 millimeters, airspeed of 220 meters per second, and external temperature of minus 20 degrees Celsius, the lookup table returns a target duty cycle of 65%, and the controller adjusts the PWM output accordingly, reducing the duty cycle from the initial 95% to 65% and maintaining a switching frequency of 1 kHz. At the moment when the new duty cycle instruction takes effect, the current system timestamp is recorded, for example, 1.51 seconds, and the effective parameter group (including area identification, PWM frequency 1 kHz, and duty cycle 65%) is packaged into a data structure. This data structure containing the timestamp and the new parameter group marks the beginning of a new heating phase, generating a phase transition maintenance pulse sequence.

[0055] The acquisition step of the area melting completion identifier is: According to the phase transition maintenance pulse sequence, the temperature-time data is processed in a fixed length time window in sequence during the pulse application period, and the real-time temperature change slope is updated; According to the real-time temperature change slope, continuously make monotonic discrimination with the melting end slope threshold and track the first jump point after the near-zero interval, and when the real-time temperature change slope is greater than the melting end slope threshold, write the current timestamp and current temperature to form the area melting completion identifier.

[0056] Specifically, according to the phase transition maintenance pulse sequence, during the entire period of its continuous application to the heating element, the data processing flow runs uninterruptedly, collecting data from the temperature sensor at a fixed frequency of 100 Hz, and applying the previously established abnormal fluctuation elimination and median filtering algorithm to preprocess the raw data, ensuring the quality of the data stream, the processed temperature time data points are sequentially sent to a first-in-first-out (FIFO) buffer with a length of 15 data points, which constitutes a sliding fixed-length time window in time, every time a new data point enters the buffer, the oldest data point is removed, keeping the window always containing continuous temperature time data within the last 0.15 seconds, for every update of the buffer, i.e. every 0.01 seconds, the linear regression slope calculation formula is called to calculate the 15 data points in the window, obtaining a new slope value, this process continues uninterruptedly, generating a high refresh rate slope data stream that faithfully reflects the subtle dynamics of temperature change during the phase transition maintenance stage, since the main energy in this stage is used for ice melting rather than temperature rising, the calculated slope value will theoretically fluctuate within a small positive value interval close to zero, for example, between -0.5 and 0.5 degrees Celsius per second, which is in contrast to the high positive slope in the initial heating stage, updating the real-time temperature change slope.

[0057] According to the continuously updated real-time temperature change slope data stream, a preset melting end slope threshold is compared, which is set based on the following empirical method: in a large number of ground deicing experiments, a high-speed camera is used to record the moment when the ice layer completely disappears, and all real-time temperature change slope values calculated within 1 second after that moment are extracted, a statistical analysis is performed on the set of slope values representing the temperature rise of liquid water, the mean and standard deviation are calculated, and finally the melting end slope threshold is set to the mean plus twice the standard deviation, this can stably distinguish the temperature rise signal after melting from the near-zero slope fluctuation during phase transition, for example, if the mean is 3.5 degrees Celsius per second and the standard deviation is 0.75 degrees Celsius per second, the threshold is set to The determination logic first confirms that the stable phase transition plateau has been entered, i.e. the real-time temperature change slope has been continuously monitored within the near-zero interval (e.g. -0.5 to 0.5 degrees Celsius per second) for at least 1 second, once the plateau is confirmed, the determination logic starts tracking whether the slope jumps, when the first detection of 3 (a tunable parameter to prevent single-point noise false triggering) consecutive real-time temperature change slopes all greater than 5.0 degrees Celsius per second, the determination condition is met, at this time, the precise timestamp and temperature value corresponding to the first point in the 3 points are captured, for example (15.23 seconds, 1.2 degrees Celsius), and these two values are encapsulated to form the regional melting completion identifier.

[0058] The obtaining step of the updated deicing area priority sequence is: According to the area melting completion identifier, a stop supply instruction is issued to the heating element of the current area, the pulse energy output is turned off, the supply current value and the duty cycle value are read to confirm zero, the stop supply effective timestamp, the last pulse parameter and the current area identifier are recorded, and the stop supply confirmation state is generated; According to the stop supply confirmation state, the current area identifier is searched and single deleted in the deicing area priority sequence, rearranged in descending order of the remaining area aerodynamic penalty index, the rearrangement timestamp and the sequence checksum are recorded, and the updated deicing area priority sequence is generated.

[0059] Specifically, according to the area melting completion identifier generated in the previous step, the stop supply program is triggered, a hardware stop supply instruction is sent to the power controller of the current heating area (for example, B1 area), the instruction is transmitted through a dedicated digital I / O line or CAN bus message, and the pulse width modulation (PWM) output pin of the power controller is forced to enter a high resistance state or a low voltage state, thereby completely shutting down the pulse energy output. Within 10 milliseconds after sending the instruction, a feedback verification loop is executed, first, the current value of the supply circuit is read through the Hall current sensor connected in series with the heating element, second, the current duty cycle setting value is read back from the state register of the power controller, and then the two read-back values are compared with the preset zero value threshold. The zero value threshold of the current is set to 0.05 amperes, and the zero value threshold of the duty cycle is set to 0.1%. Only when the supply current value is less than 0.05 amperes and the duty cycle value is less than 0.1% can the energy supply be completely interrupted, and the confirmation is successful. After confirmation, the current system timestamp is immediately recorded as the stop supply effective timestamp, for example, 15.25 seconds, and the last pulse parameter (for example, PWM frequency 1 kHz, duty cycle 65%) before the stop supply and the current processing area identifier (B1 area) are packaged together to generate a stop supply confirmation state.

[0060] According to the de-icing confirmation state generated in the previous link, the task scheduler is activated, first access the de-icing area priority sequence currently stored in the memory, which is an ordered list containing all the region identifiers to be processed and their corresponding aerodynamic penalty indexes, for example {(A1 region: 0.215), (A2 region: 0.211), (C1 region: 0.180)}, then the program uses the current region identifier (B1 region) recorded in the de-icing confirmation state as the search key to find the matching entry in the sequence, after finding the matching entry, perform a one-time deletion operation to remove the entry from the sequence, then, according to the aerodynamic penalty indexes of the remaining region list, perform a one-time descending arrangement operation again, since the previous sorting may trigger the secondary sorting rule based on the angle of attack or airspeed, this time the rearrangement will be strictly based on the penalty index value, without considering the secondary rule, unless the difference is less than the preset criterion again, after the rearrangement is completed, record the system timestamp at the time of operation completion, and perform a cyclic redundancy check (CRC32) on all region identifiers in the newly generated sequence, calculate a sequence checksum, which will be used for subsequent data transmission and storage integrity verification, for example, the rearranged sequence is {(A1 region: 0.215), (A2 region: 0.211), (C1 region: 0.180)}, record the rearrangement timestamp 15.28 seconds and the calculated checksum 0xDEADBEEF, generate the updated de-icing area priority sequence.

[0061] The next step of obtaining the heating target instruction is: According to the updated de-icing area priority sequence, select the highest ranked region identifier and fill in the initial heating pulse parameters, temperature sampling period and phase change start slope threshold, and perform the S2 and S2 steps in a loop to form the next cycle heating target instruction.

[0062] Specifically, according to the newly generated updated deicing area priority sequence, the control program automatically selects the area with the highest current ranking as the target of the next heating cycle, for example, selects A1 area from the sequence {(A1 area: 0.215), (A2 area: 0.211), (C1 area: 0.180)}, then the program fills a set of standard start-up parameters for this new target area from a pre-set parameter configuration library, this set of parameters includes initial heating pulse parameters for starting rapid heating (for example, PWM frequency 1 kHz, duty cycle 95%, duration 3 seconds), temperature sampling period of the data acquisition system (for example, 10 milliseconds, corresponding to a sampling rate of 100 Hz) and phase change start-up slope threshold for judging the start of ice crystal melting (for example, 0.55 degrees Celsius per second), these parameters are packaged into a structured instruction set, then the control logic enters a loop execution state, takes this newly generated instruction set as input, and restarts the S2 step, that is, applies an initial heating pulse to the heating elements of the A1 area, and starts collecting and analyzing temperature data, calculates the real-time temperature change slope, then enters the S3 step, judges the phase change start and switches to the phase change maintenance pulse, the entire deicing process restarts in this way on the new target area, this process will continue to repeat until the updated deicing area priority sequence is empty, forming the next cycle heating target instruction.

[0063] The above is only a preferred embodiment of the present application, and does not limit the present application in other forms. Any skilled person in the art can modify or change the above disclosed technical content to equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made according to the technical essence of the present application to the above embodiments without departing from the technical solution content of the present application still belongs to the protection scope of the present application technical solution.

Claims

1. A method of pulse-heated aircraft de-icing, characterized in that, The method comprises the following steps: S1, collecting the ice layer thickness, ice layer type, air speed and angle of attack of each aircraft surface area, calling the inherent aerodynamic sensitivity coefficient of each area, establishing the aerodynamic penalty index of each area, arranging the aerodynamic penalty index of each area in descending order according to the numerical value, and generating a deicing area priority sequence; S2, according to the deicing area priority sequence, selecting the first area in the sequence to apply an initial heating pulse to the heating element, and continuously determining the real-time temperature change slope, when the real-time temperature change slope drops to a preset phase change starting slope threshold, an ice crystal phase change starting feature point is established; S3, according to the ice crystal phase change starting feature point, adjust the pulse parameters applied to the heating element to form a phase change maintenance pulse sequence, and continuously calculate the temperature change slope of the temperature time data during the application of the phase change maintenance pulse sequence, update the real-time temperature change slope, when the real-time temperature change slope jumps from the state of approaching zero and exceeds the melting end slope threshold, a region melting completion identifier is established; S4, according to the region melting completion identifier, stop supplying pulse energy to the heating element of the current area, and remove the processed area from the deicing area priority sequence to obtain an updated deicing area priority sequence, select the highest ranked area in the updated deicing area priority sequence as the next target, and cyclically execute steps S2 and S2 to establish a next cycle heating target instruction.

2. The method of claim 1, wherein, The deicing area priority sequence acquisition step is: Analyze each aircraft surface area identifier and collect ice layer thickness value, ice layer type identifier, air speed value and angle of attack value one by one, call the aerodynamic sensitivity coefficient value matched with the area number, and introduce the area characteristic length value and stall angle of attack value, normalize to generate a region parameter list; According to the region parameter list, the region aerodynamic penalty index is calculated; According to the region aerodynamic penalty index, the region aerodynamic penalty index is arranged in descending order from high to low, if the absolute value of the difference between the region aerodynamic penalty indexes of adjacent regions is less than a preset criterion, the absolute value of the angle of attack is sorted, if it is still the same, the air speed is sorted, and a deicing area priority sequence is formed.

3. The method of claim 1, wherein, The real-time temperature change slope acquisition step is: According to the deicing area priority sequence, locate the first area and apply an initial heating pulse to the area heating element, continuously read the timestamp and temperature value and eliminate abnormal fluctuation points, align the data in timestamp order to obtain a temperature time data sequence; According to the temperature time data sequence, the real-time temperature change slope is calculated.

4. The method of claim 1, wherein, The ice crystal phase change starting feature point acquisition step is: According to the real-time temperature change slope, compare it with the phase change starting slope threshold point by point, when the real-time temperature change slope is less than or equal to the phase change starting slope threshold for the first time, read the corresponding temperature value and timestamp and combine them into a fixed value to generate an ice crystal phase change starting feature point.

5. The method of claim 1, wherein, The phase change maintenance pulse sequence acquisition step is: According to the ice crystal phase change starting feature point, a phase change maintenance switching time is located, a short pulse is terminated and a medium power parameter is loaded, a duty cycle is up-regulated from an initial duty cycle to a target duty cycle, and a valid timestamp and a parameter group are recorded to generate a phase change maintenance pulse sequence.

6. The method of claim 1, wherein, The obtaining step of the region melting completion identifier is: According to the phase change maintenance pulse sequence, temperature-time data is sequentially processed in a fixed length time window during pulse application, and a real-time temperature change slope is updated; According to the real-time temperature change slope, a monotonic discrimination with a melting end slope threshold is continuously performed, and a first jump point after a near-zero interval is tracked, and when the real-time temperature change slope is greater than the melting end slope threshold, a current timestamp and a current temperature are written to form the region melting completion identifier.

7. The method of claim 1, wherein, The obtaining step of the updated deicing region priority sequence is: According to the region melting completion identifier, a stop supply instruction is issued to a heating element of a current region, pulse energy output is turned off, a supply current value and a duty cycle value are read to confirm zero, a stop supply valid timestamp, a last pulse parameter and a current region identifier are recorded, and a stop supply confirmation state is generated; According to the stop supply confirmation state, the current region identifier is searched in the deicing region priority sequence and is deleted once, the remaining regions are rearranged in descending order of region aerodynamic penalty indexes, a rearrangement timestamp and a sequence checksum are recorded, and an updated deicing region priority sequence is generated.

8. The method of claim 1, wherein, The obtaining step of the next cycle heating target instruction is: According to the updated deicing region priority sequence, a region identifier with the highest ranking is selected and filled with an initial heating pulse parameter, a temperature sampling period and a phase change starting slope threshold, and the steps S2 and S2 are executed in a loop to form a next cycle heating target instruction.

9. A method of de-icing an aircraft by pulsed heating according to any of claims 1 to 8, characterized in that It comprises: A perception evaluation module is configured to collect ice layer thickness, ice layer type, airspeed and angle of attack of each aircraft surface region, retrieve inherent aerodynamic sensitivity coefficients of each region, establish region aerodynamic penalty indexes, arrange the region aerodynamic penalty indexes in descending order according to their values, and generate a deicing region priority sequence; A heating start module is configured to select a heating element of a first region in the sequence according to the deicing region priority sequence to apply an initial heating pulse, and continuously determine a real-time temperature change slope, and when the real-time temperature change slope decreases to a preset phase change starting slope threshold, an ice crystal phase change starting feature point is established; A phase change maintenance module is configured to adjust pulse parameters applied to the heating element according to the ice crystal phase change starting feature point to form a phase change maintenance pulse sequence, and continuously perform temperature change slope calculation on temperature-time data during application of the phase change maintenance pulse sequence to update a real-time temperature change slope, and when the real-time temperature change slope jumps from a state close to zero and exceeds a melting end slope threshold, a region melting completion identifier is established. A cycle control module is configured to, according to the region melting completion identifier, stop supplying the pulsed energy to the heating element of the current region, remove the processed region from the deicing region priority sequence, obtain an updated deicing region priority sequence, select a region with the highest priority in the updated deicing region priority sequence as a next target, and establish a next cycle heating target instruction.

Images