A method and system for pulse heating aircraft de-icing

Abstract

The present application relates to the technical field of aircraft electric heating deicing, in particular to a kind of pulse heating aircraft deicing method and system, comprising the following steps: collecting the ice layer thickness of each aircraft surface area, ice layer type, air speed and angle of attack, call each area inherent aerodynamic sensitivity coefficient, establish each area aerodynamic penalty index, according to the numerical value of each area aerodynamic penalty index descending arrangement, generate deicing area priority sequence.The present application is through the joint evaluation of ice layer thickness, ice layer type, air speed and angle of attack, optimizes heating resource allocation order, avoids energy waste, and in local heating execution phase, temperature change slope real-time monitoring is used instead of fixed time control, can dynamically respond to ice layer actual thermal response behavior, through phase change start slope threshold value identification ice crystal initial phase change process, and then switch to the pulse control mode of medium power, form steady-state regulation heating path, strengthen phase change control continuity.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aircraft electric heating deicing, and particularly relates 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, destroy 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 the processing order of the areas, and prone to priority configuration errors, thereby causing a reaction lag in critical 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:

[0005] 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 aerodynamic penalty index values of each area in descending order, and generating a deicing area priority sequence;

[0006] 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;

[0007] 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;

[0008] 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 steps S2 to S4 to establish the next cycle heating target instruction.

[0009] Preferably, the step of obtaining the deicing region priority sequence is:

[0010] The region parameter list is generated by parsing each aircraft surface region identifier, collecting ice layer thickness values, ice layer type identifiers, airspeed values, and angle of attack values one by one, calling aerodynamic sensitivity coefficient values matched with region numbers, and introducing region characteristic length values and stall angle of attack values, and normalizing.

[0011] According to the region parameter list, the region aerodynamic penalty index is calculated.

[0012] According to the region aerodynamic penalty index, the region aerodynamic penalty index is arranged in descending order from high to low, and 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, and if it is still the same, the airspeed is sorted, forming the deicing region priority sequence.

[0013] Preferably, the step of obtaining the real-time temperature change slope is:

[0014] According to the deicing region priority sequence, the first region is located and an initial heating pulse is applied to the region heating element. The time stamp and temperature value are continuously read and abnormal fluctuation points are removed. The data is aligned in time stamp order to obtain a temperature time data sequence.

[0015] According to the temperature time data sequence, the real-time temperature change slope is calculated.

[0016] Preferably, the step of obtaining the ice crystal phase transition starting feature point is:

[0017] According to the real-time temperature change slope, the real-time temperature change slope is compared with the phase transition starting slope threshold value point by point. When the real-time temperature change slope is less than or equal to the phase transition starting slope threshold value for the first time, the corresponding temperature value and time stamp are read and combined into a pair of fixed values to generate the ice crystal phase transition starting feature point.

[0018] Preferably, the step of obtaining the phase transition maintenance pulse sequence is:

[0019] According to the ice crystal phase transition starting feature point, a phase transition maintenance switching moment 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, an effective time stamp and a parameter group are recorded, and a phase transition maintenance pulse sequence is generated.

[0020] Preferably, the obtaining step of the regional melting completion identifier is:

[0021] According to the phase transition 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;

[0022] According to the real-time temperature change slope, a monotonic discrimination is continuously performed with a melting end slope threshold value, 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 value, a current time stamp and a current temperature are written to form the regional melting completion identifier.

[0023] Preferably, the obtaining step of the updated deicing region priority sequence is:

[0024] According to the regional melting completion identifier, a stop supply instruction is issued to a heating element of a 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 time stamp, the last pulse parameter and the current region identifier are recorded, and a stop supply confirmation state is generated;

[0025] 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 the regional aerodynamic penalty index, the rearrangement time stamp and the sequence checksum are recorded, and the updated deicing region priority sequence is generated.

[0026] Preferably, the obtaining step of the next cycle heating target instruction is:

[0027] According to the updated deicing region priority sequence, the region identifier with the highest ranking is selected and filled with the initial heating pulse parameter, the temperature sampling period and the phase transition starting slope threshold value, and the steps S2 to S4 are executed in a loop to form the next cycle heating target instruction.

[0028] The application also provides an aircraft deicing system, comprising:

[0029] The 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 of the regions, and generate a deicing region priority sequence.

[0030] The heating starting module is used for selecting a heating element of a first area in a sequence according to the deicing area priority sequence to apply an initial heating pulse, and continuously determining a real-time temperature change slope, and when the real-time temperature change slope falls to a preset phase change starting slope threshold, an ice crystal phase change starting feature point is established;

[0031] The phase change maintaining module is used for adjusting a pulse parameter applied to the heating element according to the ice crystal phase change starting feature point to form a phase change maintaining pulse sequence, continuously performing temperature change slope operation on the temperature time data during application of the phase change maintaining 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, a region melting completion identifier is established.

[0032] The cycle control module is used for stopping supply of pulse energy to the heating element of the current area according to the region melting completion identifier, removing the processed area from the deicing area priority sequence to obtain an updated deicing area priority sequence, selecting a region with the highest ranking in the updated deicing area priority sequence as a next target, and establishing a next cycle heating target instruction.

[0033] Compared with the prior art, the application has the following advantages and positive effects:

[0034] The application optimizes a heating resource allocation sequence through joint evaluation of ice layer thickness, ice layer type, airspeed and angle of attack, avoids energy waste, and uses real-time temperature change slope monitoring to replace fixed time control in a local heating execution stage, can dynamically respond to actual thermal response behavior of the ice layer, identifies an ice crystal initial phase change process through a phase change starting slope threshold, and then switches to a pulse control mode with medium power to form a steady-state regulated heating path, strengthens phase change control continuity, continuously samples a temperature change trend in the phase change process, and identifies a melting completion node based on slope jump, updates a deicing area priority sequence in a timely manner based on a region melting completion identifier, realizes real-time synchronization of processing progress, improves real-time performance of deicing response, and avoids increased power consumption caused by lagging or redundant heating strategies. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 The application is illustrated by the following specific embodiments. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical scheme and advantages of the application clearer, the application is further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application.

[0037] Please refer to Figure 1The application provides a technical scheme, a method for deicing an airplane by pulse heating, comprising the following steps:

[0038] S1, collecting the ice layer thickness, ice layer type, air speed and angle of attack of each airplane 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;

[0039] 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;

[0040] 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, 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 region melting completion identifier;

[0041] S4, according to the region melting completion identifier, stopping the supply of pulse energy to the heating element of the current area, removing the processed area in the deicing area priority sequence, obtaining an updated deicing area priority sequence, selecting the area with the highest ranking in the updated deicing area priority sequence as the next target, and cyclically executing steps S2 to S4 to establish a next cycle heating target instruction.

[0042] The deicing area priority sequence acquisition step is:

[0043] The ice layer thickness value, ice layer type identifier, air speed value and angle of attack value of each airplane surface area identifier are collected one by one, the aerodynamic sensitivity coefficient value matched with the area number is called, and the area characteristic length value and stall angle of attack value are introduced, normalized, and a region parameter list is generated;

[0044] According to the region parameter list, the region aerodynamic penalty index is calculated, and the calculation formula is:

[0045] ;

[0046] Among them, 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, is the ice layer thickness effect index, is the air density value, is the regional airspeed value, is the reference dynamic pressure value, is the regional angle of attack absolute value, is the regional stall angle of attack value, , , is the weight coefficient determined by experiment;

[0047] According to the regional aerodynamic penalty index, the regional aerodynamic penalty index is arranged in descending order from high to low, and if the absolute value of the difference of the regional aerodynamic penalty index of adjacent regions is less than a preset criterion, the size of the angle of attack absolute value is sorted, and if it is still the same, the size of the airspeed is sorted, forming a deicing region priority sequence.

[0048] Specifically, according to the analyzed each aircraft surface region identifier, such as wing leading edge A1 region, wing leading edge A2 region, horizontal tail B1 region, vertical tail C1 region, etc., start the multi-channel sensor data acquisition process, for each identifier region, the ultrasonic sensor array deployed performs pulse echo detection at a frequency of 50 hertz, by calculating the time difference of ultrasonic reflection at the ice layer and skin interface, combined with the sound speed in ice (about 3800 meters / second) to convert the ice layer thickness value, the synchronous deployment of the optical ice layer detector analyzes the reflection spectrum characteristics, identifies the ice layer type as clear ice, frost ice or mixed ice, and records this classification information, at the same time, read the airspeed and angle of attack values from the aircraft ARINC 429 data bus in real time, these two parameters directly reflect the current flight state of the aircraft, then, access the aerodynamic characteristic database stored in the non-volatile memory of the onboard computer, according to the current processed region identifier (such as A1 region) 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, normalize all collected and retrieved dynamic and static values, use the minimum-maximum normalization method to linearly map each parameter value to the interval, the normalization formula is , where is the original parameter value, and are the historical minimum and maximum values 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 / second, 250 m / s, finally, each zone identifier is integrated with all its corresponding raw and normalized values into a structured data record, and all zone data records together form the zone parameter list.

[0049] In the zone aerodynamic penalty index formula, the physical properties of icing (thickness, type), the geometric and aerodynamic characteristics of the zone (sensitivity, characteristic length, stall angle of attack), and the current flight state (airspeed, angle of attack) are integrated into a unified evaluation framework. Not only does it consider the direct impact of ice thickness , but also simulates the nonlinear disturbance effect of ice shape on airflow through the ratio of characteristic length and index , and introduces the coupling of dynamic pressure term and angle of attack term , which enables the penalty index to dynamically reflect the rapid increase in risk when the aircraft approaches the stall boundary, thus achieving a more intelligent and risk-oriented allocation of deicing resources.

[0050] is the zone aerodynamic sensitivity coefficient value, which represents the impact of icing on the specific area of the aircraft surface 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 aircraft design phase. In the simulation model, a standard-shaped simulated ice shape is applied to each key area 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 area is compared with the most severe performance degradation among all areas, and the normalized value is obtained, which is , a value between 0 and 1, the higher the value, the more sensitive the area, for example, the center area of the wing leading edge is most sensitive to lift, its value is usually close to 1, while the value of the fuselage belly area is lower. A specific example is: for a wing tip area, CFD simulation shows that icing leads to a 5% decrease in total lift coefficient and an 8% increase in total drag coefficient, while the most sensitive area of the aircraft (the 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 respectively, then the comprehensive influence degree of the wing tip area is , the influence degree of the most sensitive area is , then the value of the wing tip area is , in this example, the of a key area of the wing leading edge is .

[0051] is a dimensionless coefficient quantifying the difference of aerodynamic performance influenced by different ice layer types. Since the non-numeric ice layer type identification (such as clear ice, rime ice) cannot be directly substituted into the formula for calculation, a quantitative mapping standard needs to be established. The establishment of this standard is based on icing wind tunnel experimental data. Experiments show that the disturbance of smooth and dense clear ice to airflow is much greater than that of rough and low-density rime ice. Therefore, by analyzing the influence curve of different ice types on the lift-drag ratio characteristics of airfoils in a large amount of experimental data, a baseline (usually rime 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 the lift loss caused by clear ice is 1.4 times that of rime ice, so its value can be set to 1.4. An example of obtaining is: through the onboard optical sensor, the current regional ice layer is identified as "clear ice", and the mapping table is consulted to obtain the value of 1.4.

[0052] is the regional ice layer thickness value, with the unit of meters (m). This value 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 the measurement data to the deicing control computer. The computer performs sliding average filtering on the continuous readings to eliminate measurement noise of individual pulses, obtaining a stable and representative thickness value. For example, in a certain sampling period, the sensor continuously measures the thickness as 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.

[0053] is the regional characteristic length value, with the unit of meters (m). This is a fixed parameter related to the geometric shape of the aircraft, representing the macroscopic size of a specific region. For airfoil surfaces such as wings or tail wings, the local chord length is usually taken. For other components such as the nose, the radius of curvature or representative length is taken. This parameter is determined during the aircraft design phase and is pre-stored in the onboard database. It can be called through the region identifier. For example, for the wing root region of a medium-sized passenger aircraft, the characteristic length may be m.

[0054] 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 influence 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, the lift loss data points of a certain airfoil at different ice thicknesses are fitted, if it is found that the lift loss is proportional to the 1.2 power of , then it is determined that is .

[0055] is the air density value, with the unit of kilogram per cubic meter (kg / m 3 ), which is calculated by the atmospheric data computer (ADC) according to the static pressure and total temperature at the current flight altitude, in accordance with the international standard atmospheric model (ISA), for example, at 8000 meters of cruising altitude, the air density under standard atmospheric conditions is about kg / m 3 .

[0056] is the regional airspeed value, with the unit of meters per second (m / s), which is also provided by the atmospheric data computer, representing the speed of the aircraft relative to the surrounding air, using the true airspeed (TAS) provided by the ADC as the unified , for example, if the current true airspeed of the aircraft is 220 m / s, then m / s.

[0057] is the reference dynamic pressure value, with the unit of pascal (Pa), which is a reference value for normalizing the current dynamic pressure, so that the penalty index has comparability under different flight conditions, usually the dynamic pressure of the aircraft flying at the standard sea level (about kg / m 3 ) at its maximum cruising speed (for example m / s) is selected as the reference value, and the calculation formula is: , and the setting process is: Pa.

[0058] is the absolute value of the regional angle of attack, with the unit of degree (°), provided by the flight control system or the atmospheric data computer, representing the angle between the chord line of the wing and the direction of the far front flow, for example, the current angle of attack of the aircraft is .

[0059] is the regional stall angle of attack value, with the unit of degree (°), which is the angle of attack at which the specific airfoil region can achieve the maximum lift coefficient under the ice-free state, and it is an inherent aerodynamic characteristic parameter, which is pre-stored in the onboard database, and beyond this angle of attack, the lift will decrease sharply, resulting in stall, for example, for the selected airfoil of the region, the stall angle of attack under the clean state is For .

[0060] , , are experimentally determined weight coefficients, these dimensionless coefficients are used to adjust the relative importance of different terms in the formula, penalty weight that adjusts the ice layer geometry (thickness), penalty weight that adjusts the current aerodynamic load, and then specifically adjust the risk amplification effect when the aircraft approaches stall condition, their settings are not subjective set, but through an optimization process, the specific steps are: collect a large number of flight data or high-fidelity simulation data, these data contain various icing conditions and their corresponding actual flight risk level (rated by test pilots or quantified by performance degradation), then, with the objective function of minimizing the difference between the formula calculated ranking and the actual risk ranking, use genetic algorithm or particle swarm optimization algorithm for iterative optimization, finally determine a set of optimal weight coefficients, the weight values determined through this process are: , , .

[0061] According to the above parameters for calculation:

[0062] The obtained numerical value is substituted into the formula:

[0063] .

[0064] .

[0065] m.

[0066] m.

[0067] .

[0068] kg / m 3 .

[0069] m / s.

[0070] Pa.

[0071] ;

[0072] ;

[0073] ;

[0074] ;

[0075] ;

[0076] Calculation process:

[0077] ;

[0078] ;

[0079] ;

[0080] ;

[0081] ;

[0082] ;

[0083] ;

[0084] ;

[0085] ;

[0086] ;

[0087] ;

[0088] ;

[0089] This result shows that the current wing leading edge key area aerodynamic penalty index calculated value is 0.21536, after obtaining all icing areas values, this value will be used for sorting, the higher the value of the area, the greater the threat of the icing state to flight safety, should be given a higher priority for deicing, for example, if another area (such as the horizontal tail) calculated value is 0.35, it means that the horizontal tail is more urgent than this wing area for deicing, and the horizontal tail will be handled first.

[0090] According to the calculated aerodynamic penalty index of each region, all regions and their corresponding penalty index values are taken as a set, for example, {A1 region: 0.215, A2 region: 0.211, B1 region: 0.350, C1 region: 0.180}, first, the first level sorting rule is executed, that is, strictly descending according to the numerical value of the penalty index, and a preliminary sorting {B1 region: 0.350, A1 region: 0.215, A2 region: 0.211, C1 region: 0.180} is obtained, then, the second level sorting rule is started, that is, the penalty indexes of adjacent regions are compared to check whether the absolute value of their difference is less than a 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 switching caused by slight numerical fluctuations, and the setting process is as follows: the average of 1.5 times the standard deviation of the penalty index values of each region in the stable flight stage in the last 100 flights is taken as the criterion, for example, the average standard deviation is calculated as 0.008, and the criterion is set as In the above sequence, the program checks the difference between A1 region and A2 region, 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 A1 region and A2 region, for example, the angle of attack of A1 region is 4.5 degrees and the angle of attack of A2 region 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 region with larger angle of attack should be promoted, therefore the sequence is adjusted to {B1 region: 0.350, A2 region: 0.211, A1 region: 0.215, C1 region: 0.180}, if the absolute values of the angles of attack of the two regions are also the same, then the third level sorting rule is entered, that is, the airspeed of the two regions is compared, the region with higher airspeed is given higher priority because the icing growth rate may be faster or the energy demand is larger, after all the comparisons and adjustments are completed, a stable and multi-factor considering deicing region priority sequence is finally obtained.

[0091] The steps for obtaining the real-time temperature change slope are as follows:

[0092] According to the deicing region priority sequence, the first region is located and an initial heating pulse is applied to the region 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;

[0093] According to the temperature time data sequence, the real-time temperature change slope is calculated, and the calculation formula is:

[0094] ;

[0095] Wherein, is the real-time temperature change slope at the kth time point, is the size of the sliding window, is the current index, is the ith timestamp, is the ith temperature value.

[0096] Specifically, according to the previously obtained sequence of deicing area priorities, the highest ranked area in the sequence is first located as the current processing target, for example, the B1 area (horizontal tail fin), and then an instruction is issued to the power controller of the heating element array integrated to the area to apply a preset initial heating pulse, the pulse being characterized by high power and short time, with 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, an abnormal fluctuation elimination logic is performed on each new temperature data point collected, which first judges 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, so 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, forming a continuous and clean temperature time data sequence.

[0097] 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 short period of time. 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. ), it utilizes the information of data points in the window, smoothing 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 raising the temperature of the substance (sensible heat) to changing the phase of the substance (latent heat), resulting in a small but clear inflection point in the temperature rise rate.

[0098] The sliding window size, a dimensionless positive integer, determines the number of data points used to calculate the slope. Its selection is a crucial tuning process, requiring a balance between response speed and signal smoothness; a window that is too small... A value that is too large will cause the calculated slope to be sensitive to noise and fluctuate wildly, while an excessively large value will cause the slope to be sensitive to noise and fluctuate wildly. 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.

[0099] 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.

[0100] 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.

[0101] 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.

[0102] Calculations based on parameters:

[0103] 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:

[0104] Timestamp (Unit: s):

[0105] {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}.

[0106] Temperature (in °C):

[0107] {-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}.

[0108] Calculate the numerator:

[0109] ;

[0110] ;

[0111] ;

[0112] ;

[0113] ;

[0114] Substitute these values into the formula:

[0115] Molecular calculation:

[0116] ;

[0117] ;

[0118] Denominator calculation:

[0119] ;

[0120] ;

[0121] Slope calculation:

[0122] ;

[0123] The results show that at time point 0.20, the temperature rise rate obtained by linear regression analysis of the data from the previous 0.15 seconds was approximately 36.26 degrees Celsius per second. This value reflects that in the initial heating stage, the ice layer and aircraft skin are rapidly absorbing heat and the temperature is rising steadily. This slope value will be used as a continuous output for real-time comparison with the phase change initiation slope threshold of the next stage. A high and stable positive slope value (such as 36.26) confirms that the effective sensible heat heating stage has not yet begun and the phase change process has not yet started. When the subsequently calculated slope value begins to decrease continuously and approaches a preset low threshold, it indicates that the melting process is about to begin.

[0124] The steps for obtaining the initiation feature points of ice crystal phase transition are as follows:

[0125] Based on the real-time temperature change slope, the slope is compared point by point with the phase change initiation slope threshold. When the real-time temperature change slope is first detected to be less than or equal to the phase change initiation slope threshold, the corresponding temperature value and timestamp are read and combined into a pair of fixed values ​​to generate the ice crystal phase change initiation feature point.

[0126] 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" less than or equal to 0.55 and the phase change detection trigger is false, once this condition is first met, the program immediately sets the phase change detection trigger flag to true and simultaneously reads the time stamp and temperature value corresponding to this point from the current temperature time data sequence, for example, when it is calculated that the condition is met, the program reads the time stamp as 1.50 seconds and the temperature as -0.05 degrees Celsius, and then combines these two values into a fixed binary tuple (1.50, -0.05) to generate an ice crystal phase change start feature point.

[0127] The acquisition step of the phase change maintenance pulse sequence is:

[0128] According to the ice crystal phase change start feature point, the phase change maintenance switching time is located, the short-time pulse is terminated and the medium power parameters are loaded, the duty cycle is up-regulated from the initial duty cycle to the target duty cycle, the effective time stamp and parameter group are recorded, and the phase change maintenance pulse sequence is generated.

[0129] Specifically, the time information contained in the ice crystal phase change start feature point obtained in the previous step, such as 1.50 seconds, is directly used as the phase change maintenance switching time, and the control system immediately issues an instruction to the power controller of the current regional heating element to terminate the previously applied high-power short-time pulse, and the specific operation is to set the duty cycle of the pulse width modulation (PWM) signal to 0% instantaneously, cutting off the energy supply, and then loading a set of pre-set medium power parameters, the core of which is the target duty cycle, the value of which is not fixed but determined dynamically through a three-dimensional lookup table, which is pre-stored in the on-board computer memory, taking the ice layer thickness of the current region, the true 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%, the controller immediately adjusts the PWM output, down-regulating the duty cycle from the initial 95% to 65%, and maintains a switching frequency of 1 kilohertz, at the moment when the new duty cycle instruction takes effect, the current system time stamp, for example, 1.51 seconds, is recorded, and the effective parameter group (including region identification, PWM frequency 1 kilohertz, duty cycle 65%) is packaged into a data structure, this data structure containing the time stamp and the new parameter group marks the beginning of a new heating phase, generating the phase change maintenance pulse sequence.

[0130] The acquisition step of the regional melting completion identifier is:

[0131] According to the phase change maintaining pulse sequence, the temperature-time data is sequentially processed in a fixed length time window during the pulse application, and the real-time temperature change slope is updated;

[0132] According to the real-time temperature change slope, the first jump point after the near-zero interval is continuously and monotonously discriminated and tracked with the melting end slope threshold value, and when the real-time temperature change slope is greater than the melting end slope threshold value, the current timestamp and the current temperature are written, forming a regional melting completion identifier.

[0133] Specifically, according to the phase change maintaining pulse sequence, during the entire period of its continuous application to the heating element, the data processing flow runs uninterruptedly, collects data from the temperature sensor at a fixed frequency of 100 Hz, and applies the previously established abnormal fluctuation elimination and median filtering algorithm to the raw data for preprocessing, 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. This buffer 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 each update of the buffer, i.e., every 0.01 seconds, a linear regression slope calculation formula is called to calculate the 15 data points in the window, obtaining a new slope value. This process continues continuously, generating a high refresh rate slope data stream that faithfully reflects the subtle dynamics of temperature change during the phase change maintaining stage. Since the main energy in this stage is used for ice melting rather than temperature rise, 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. The real-time temperature change slope is updated.

[0134] According to the continuously updated real-time temperature change slope data stream, a preset melting end slope threshold value is compared. The setting of this threshold value is 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. Statistical analysis is performed on this set of slope values representing the temperature rise of liquid water, and the mean and standard deviation are calculated. Finally, the melting end slope threshold value 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 the phase change process. 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 value is set to Celsius per second, the determination logic will first confirm that the stable phase transition plateau has been entered, that is, the real-time temperature change slope has been continuously monitored in the near-zero interval (for example, -0.5 to 0.5 Celsius per second) for at least 1 second, and once it is confirmed that the plateau has been entered, the determination logic begins to track whether the slope jumps, and when the first three (a tunable parameter to prevent single-point noise false triggering) consecutive real-time temperature change slopes are all greater than 5.0 Celsius per second, the determination condition is satisfied, at this time, the precise timestamp and temperature value corresponding to the first of the three points are immediately captured, for example, (15.23 seconds, 1.2 Celsius), and this pair of values is encapsulated to form a region melting completion identifier.

[0135] The obtaining step of the updated deicing region priority sequence is:

[0136] 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 to confirm 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;

[0137] According to the stop supply confirmation state, the current region identifier is searched and single deleted in the deicing region priority sequence, rearranged in descending order of the remaining region aerodynamic penalty index, the rearrangement timestamp and the sequence checksum are recorded, and the updated deicing region priority sequence is generated.

[0138] Specifically, according to the region melting completion identifier generated in the previous step, a stop supply program is triggered, a hardware stop supply instruction is sent to the power controller of the current region being heated (for example, B1 region), 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 turning off 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%, it is confirmed that the energy supply has been 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 region identifier (B1 region) are packaged together to generate a stop supply confirmation state.

[0139] According to the de-icing confirmation state generated in the previous step, the task scheduler is activated, first access the de-icing area priority sequence currently stored in 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 recorded in the de-icing confirmation state (B1 region) 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.

[0140] The next step of obtaining the heating target instruction is:

[0141] 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 steps S2 to S4 in a loop to form the next cycle heating target instruction.

[0142] 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.

[0143] 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 to S4 to establish the 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 steps S2 to S4 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