Interpolation method for searching the thermodynamic properties of aero-engine gas based on memory data

Abstract

The present invention discloses a method for searching and interpolating the thermodynamic properties of aircraft engine gas based on memory data, comprising the following steps: step 1) inputting data; and step 2) searching for the specific location of a target parameter in a table based on the input data entered in step 1). The present invention incorporates dynamic memory data of the thermodynamic properties of aircraft engine gas when searching for the target parameter, and through comparison with the memory data, quickly locates the specific location of the target parameter.

Description

Technical Field

[0001] The present invention relates to the technical field of mathematical modeling and simulation of overall performance of aircraft engines, and mainly relates to a method for searching and interpolating the thermodynamic properties of aircraft engine gas based on memory data. Background Art

[0002] Aircraft engine airborne adaptive models can be used for engine performance health monitoring and fault diagnosis, as well as for the design and verification of advanced control algorithms (such as performance optimization control and direct thrust control). Currently, most airborne adaptive models, both domestically and internationally, employ Kalman filter estimation methods. This approach is simple to implement and computationally fast, but a single Kalman filter estimator is not applicable to the full envelope of operating points. Nonlinear component-level mathematical models of aircraft engines, however, simulate performance based on the aerodynamic thermodynamics of individual components. While computationally complex, a single nonlinear component-level model can simulate all operating points within the full envelope of a real engine. With the rapid development of microcontrollers and research into methods to improve the real-time performance of nonlinear component-level models of aircraft engines, the direct use of component-level models as airborne models is becoming possible. However, numerous factors limit the application of component-level models onboard aircraft, such as the complexity of their computational methods and the time-consuming engine flow path calculations. Consequently, the real-time performance of component-level models on airborne equipment often fails to meet requirements. By improving the efficiency of interpolation calculations of the thermodynamic properties of aircraft engine gas, the single-step simulation calculation time of aircraft engine nonlinear component-level dynamic models can be greatly shortened, thereby removing obstacles for the direct use of aircraft engine component-level models as airborne models. Summary of the Invention

[0003] Purpose of the invention: In order to solve the problems existing in the above-mentioned background technology, the present invention provides a recursive search and interpolation method for the thermodynamic properties of aircraft engine gas based on memory data. Starting from reducing the time consumption of single flow path calculation of the model, the recursive search method for the thermodynamic properties of aircraft engine gas based on memory data is used to interpolate the thermodynamic properties of the gas. Compared with the binary search, the number of searches is reduced, thereby reducing the calculation time of a single flow path of the model.

[0004] Technical solution: To achieve the above purpose, the technical solution adopted by the present invention is:

[0005] A recursive search and interpolation method for aero-engine gas thermodynamic properties based on memory data includes:

[0006] Step S1: Assume that the data input of the algorithm for the kth search includes the target data d k , memory data and target data table S{a1,a2,a3,...,a n}, where the memory data includes the target data d of the previous search step k-1And the specific location ID of the table where the target data of the previous search step is located k-1 ;

[0007] Step S2, searching for target parameters using a recursive search and interpolation method of the thermodynamic properties of the aviation engine gas based on the memory data; wherein step S2 specifically comprises the following steps:

[0008] Step S2.1: Input the target data table according to the target parameter type (total temperature, specific enthalpy, specific entropy):

[0009] S{a1,a2,a3,...,a n}

[0010] Where S is the target data table, which contains n data elements and all are real numbers. The data are arranged in ascending or descending order. This article takes ascending order as an example to illustrate the specific process.

[0011] Step S2.2: Input the current target data d to be searched k , target data d at the previous moment k-1 , calculate the difference δd between the current target data and the previous data k , where δd k d k with d k-1 The relationship between them is:

[0012]

[0013] Where, δd k =d k -d k-1 ;

[0014] Step S2.3: Determine whether the target data to be searched is out of bounds, that is, whether the target parameter to be searched is within the valid range of the data table. If it is out of bounds, set the position of the data table where the current target data is located as the minimum and maximum values. The specific expression is:

[0015]

[0016] Step S2.4: Determine the specific location ID of the target data to be searched and the target data in the previous step k-1 Whether the current target data range requirements are met. The specific expression is:

[0017]

[0018] In the formula, sort is the position id of the previous step k-1 If the current target data range requirement is not met, continue searching;

[0019] Step S2.5, segmented table lookup, specifically;

[0020] Step S2.5.1: If the difference between the current target data and the previous data is greater than or equal to 0, that is, δd k ≥0, indicating that the target data to be found is increasing compared to the target data in the previous step. The position of the current target data in the table must be greater than the position of the target data in the previous step. The specific expression is:

[0021] id k >id k-1

[0022] In this case, the location id of the target data in the previous step k-1 +1 is the starting point, and the position ID of the table where the target data to be found is searched in ascending order k .

[0023] Step S2.5.2: If the difference between the current target data and the previous data is less than 0, δd k <0, indicating that the target data to be found is decreasing compared to the target data in the previous step. The position of the current target data in the table must be smaller than the position of the target data in the previous step. The specific formula is:

[0024] id k <id k-1

[0025] In this case, the location id of the target data in the previous step k-1 -1 is the starting point, and the position ID of the table where the target data to be found is found is searched in descending order. k .

[0026] Step S2.6: Save the target data currently being searched and the position of the target data currently being searched in the table found in the above step as memory data for calculating the thermodynamic properties of the gas at the next simulation time point.

[0027] Beneficial effects:

[0028] (1) The present invention adopts a recursive search and interpolation method for the thermodynamic properties of aircraft engine gas based on memory data to calculate the thermodynamic properties of aircraft engine gas. Compared with the binary search method, it effectively reduces the number of searches and can reduce the calculation time of the nonlinear model flow path.

[0029] (2) This patent introduces memory data to save the target data d of the thermodynamic properties of the gas at the last simulation time point. k-1 The location ID in the table where the target data at the previous simulation time point is located k-1, when the target parameters at the current simulation time point are not much different from those at the previous simulation time point, quickly find the location id of the table where the target data to be searched is located k . BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 Flowchart of the calculation of the thermodynamic properties of gas of compressor components using the recursive lookup and interpolation method of the thermodynamic properties of gas of an aircraft engine based on memory data;

[0031] Figure 2 Flowchart of the recursive search and interpolation method for the thermodynamic properties of aircraft engine gas based on memory data;

[0032] Figure 3 Gas thermodynamic property calculation function test input data curve;

[0033] Figure 4 Dynamic real-time simulation test of turboshaft engine input fuel supply curve. DETAILED DESCRIPTION

[0034] The calculation of the thermodynamic properties of aviation engine gas adopts the variable specific heat method, which is specifically a high-order polynomial fitting formula. Both forward and reverse calculations are very time-consuming. In order to improve the real-time performance of the model, the usual method is to calculate several point data in advance, and then directly interpolate after looking up the table when using it. The search method usually uses a binary search. This patent discloses a new search method with fewer search times than the binary search, which improves the real-time performance of the engine model calculation.

[0035] Figure 1 A flow chart of the recursive lookup and interpolation method for aviation engine gas thermodynamic properties based on memory data for the calculation of gas thermodynamic properties of compressor components. Figure 1 The functions involved are described as follows:

[0036] (1) Calculate the specific enthalpy function from the total temperature

[0037] CalcuH(f in ,T in ,T1 in ,id Tin )

[0038] Where, f in is the inlet oil-gas ratio, T in is the total inlet temperature, T1 in is the total inlet temperature at the last simulation time point, id Tin It is the position of the inlet total temperature at the previous simulation time point in the total temperature data table.

[0039] (2) Calculate the specific entropy function from the total temperature

[0040] CalcuS(f in ,T in ,T1 in ,id Tin )

[0041] (3) Calculate the specific enthalpy function from the specific entropy

[0042] CalcuS2H(f out ,S out ,S1 out ,id S )

[0043] Where, f out is the inlet oil-gas ratio, S out is the outlet specific entropy, S1 out is the outlet entropy at the last simulation time point, id S It is the position of the outlet specific entropy at the last simulation time point in the specific entropy data table.

[0044] (4) Calculate the total temperature function from the specific enthalpy

[0045] CalcuH2T(f out ,H out ,H1 out ,id H )

[0046] Where H out is the outlet specific enthalpy, H1 out is the outlet specific enthalpy at the last simulation time point, id H It is the position of the outlet specific enthalpy at the last simulation time point in the specific enthalpy data table.

[0047] Whenever a compressor component function is called during a flow calculation, the memorized data from the previous simulation point is read before calculating the gas thermodynamic properties. At the end of the simulation, the memorized data is saved for use in the next gas thermodynamic calculation. Calling other components is similar to calling the compressor flow chart and will not be further explained.

[0048] like Figure 2 The flowchart of the method for recursively searching and interpolating the thermodynamic properties of aircraft engine gas based on memory data of the present invention is shown, which includes the following steps:

[0049] Step S1: Assume that the input of the algorithm data for the kth search includes the target data d k , memory data and target data table S{a1,a2,a3,...,a n}, where the memory data includes the target data d of the previous search step k-1 And the specific location ID of the table where the target data of the previous search step is located k-1 ;

[0050] Step S2, searching for target parameters using a recursive search and interpolation method of the thermodynamic properties of the aviation engine gas based on the memory data; wherein step S2 specifically comprises the following steps:

[0051] Step S2.1: Input the target data table according to the target parameter type (total temperature, specific enthalpy, specific entropy):

[0052] S{a1,a2,a3,...,a n}

[0053] Where S is the target data table, which contains n data elements and all are real numbers. The data are arranged in ascending or descending order. This article takes ascending order as an example to illustrate the specific process.

[0054] Step S2.2: Input the current target data d to be searched k , the target data d of the previous step k-1 , calculate the difference δd between the current target data and the target data at the previous moment k , where δd k d k with d k-1 The relationship between them is:

[0055]

[0056] Where, δd k =d k -d k-1 ;

[0057] Step S2.3: Determine whether the target data to be searched is out of bounds, that is, whether the target parameter to be searched is within the valid range of the data table. If it is out of bounds, set the position of the data table where the current target data is located as the minimum and maximum values. The specific expression is:

[0058]

[0059] Step S2.4: Determine the specific location ID of the target data to be searched and the target data in the previous step k-1 Whether the current target data range requirements are met. The specific expression is:

[0060]

[0061] In the formula, sort represents the position id of the previous step k-1 If the current target data range requirement is not met, continue searching;

[0062] Step S2.5, segmented table lookup, specifically;

[0063] Step S2.5.1: If the difference between the current target data and the previous data is greater than or equal to 0, that is, δd k ≥0, indicating that the target data to be found is increasing compared to the target data in the previous step. The position of the current target data in the table must be greater than the position of the target data in the previous step. The specific expression is:

[0064] id k >id k-1

[0065] In this case, the location id of the target data in the previous step k-1 +1 is the starting point, and the position ID of the table where the target data to be found is searched in ascending order k .

[0066] Step S2.5.2: If the difference between the current target data and the previous data is less than 0, δd k <0, indicating that the target data to be found is decreasing compared to the target data in the previous step. The position of the current target data in the table must be smaller than the position of the target data in the previous step. The specific expression is:

[0067] id k <id k-1

[0068] In this case, the location id of the target data in the previous step k-1 -1 is the starting point, and the position ID of the table where the target data to be found is found is searched in descending order. k .

[0069] Step S2.6: Save the target data currently being searched and the position of the target data currently being searched in the table found in the above step as memory data for calculating the thermodynamic properties of the gas at the next simulation time point.

[0070] Figure 3 The following table shows seven test input data curves for the gas thermodynamic calculation functions, which are the input data for the test results in Tables 1 and 3, with a total of 40,000 simulation points. The ramp-up and ramp-down of each curve are both 8,000 simulation points.

[0071] Figure 4 The figure shows the input fuel supply curve for the dynamic simulation test of the turboshaft engine, which is the input data for the test results in Tables 2 and 4. The simulation step size is 0.025, with a total of 2100 simulation points. The ramp-up and ramp-down sections of each curve have 300 simulation points.

[0072] In order to verify the effectiveness of the recursive interpolation method for finding the thermodynamic properties of aero-engine gas based on memory data designed in this invention, the method disclosed in this article and binary search are used to find various functions of the thermodynamic properties of gas. Dynamic simulation tests of a twin-spool turboshaft engine are conducted using the two methods for comparison. The test platform is:

[0073] (1) A personal PC with a 2.4GHz Intel processor, the development environment is Ubuntu 18.04, and the compiler is g++, corresponding to the data in Tables 1 and 2;

[0074] (2) The STM32MP157A development board with a 650MHz ARM Cortex-A7 processor, the development environment is Ubuntu 18.04, and the compiler is the arm-linux-gnueabihf-g++ cross-compilation tool chain, corresponding to the data in Tables 3 and 4.

[0075] Table 1 below shows a comparison of the data of the recursive search method for the thermodynamic properties of aircraft engine gas based on memory data (for the convenience of table preparation, the table is referred to as the method disclosed in this patent) and the binary search method for searching the thermodynamic properties of gas. The oil-gas ratio f is 0.01, the test platform is a personal PC with a 2.4GHz Intel processor, and the development environment is Ubuntu 18.04. Table 2 below shows a comparison of the time of the recursive search interpolation method for the thermodynamic properties of aircraft engine gas based on memory data and the binary search method when applied to the dynamic simulation of a dual-rotor turboshaft engine. The test input is Figure 4 Three sets of oil supply curves are shown, and other conditions remain unchanged at the design point; the test platform is a personal PC with a 2.4GHz Intel processor, and the development environment is Ubuntu18.04.

[0076] Table 1 Comparison of data on thermodynamic properties of gas by the public method and the binary search method

[0077]

[0078] Table 2 Comparison of the time consumption of the disclosed method and the binary search method for dynamic simulation of turboshaft engines

[0079] enter Binary search method The method disclosed in this patent Fuel supply curve 1 0.08394 0.07080 Fuel supply curve 2 0.09860 0.08504 Fuel supply curve 3 0.13190 0.11272

[0080] As shown in Table 1, the average number of searches for the search method disclosed in this article is greatly reduced compared with the binary search. On a 2.4GHz Intel processor, the time consumption of the four functions CalcuH, CalcuS, CalcuCp, and CalcuK is reduced by more than 50%, while the time consumption of the three functions CalcuS2H, CalcuH2T, and CalcuS2T is not significantly reduced. In addition, the average number of searches for the seven gas thermodynamic calculation functions is not proportional to the time consumption reduction. This is because when calculating these three functions, it is necessary to repeatedly construct a data table suitable for the oil-gas ratio based on the existing data. This process is very time-consuming, and the search is only a small part of the calculation.

[0081] As shown in Table 2, when the two methods are applied to the dynamic simulation calculation of the turboshaft engine on a 2.4 GHz Intel processor, the method disclosed in this patent reduces the time consumption by more than 13.7% compared with the binary search.

[0082] Table 3 below shows a comparison of the recursive interpolation method and the binary search method for searching for thermodynamic properties of aircraft engine gas based on memorized data. The fuel-gas ratio was 0.01, and the test platform was an STM32MP157A development board with a 650MHz ARM Cortex-A7 processor, using Ubuntu 18.04 as the development environment. Table 4 below shows a comparison of the dynamic simulation time of a twin-spool turboshaft engine using the recursive interpolation method and the binary search method based on memorized data. The test platform was an STM32MP157A development board with a 650MHz ARM Cortex-A7 processor, using Ubuntu 18.04 as the development environment.

[0083] Table 3 Comparison of gas thermodynamic data obtained by the public method and the binary search method

[0084]

[0085]

[0086] Table 4 Comparison of the time consumption of the disclosed method and the binary search method for dynamic simulation of turboshaft engines

[0087] enter Binary search method The method disclosed in this patent Fuel supply curve 1 2.11642 1.91707 Fuel supply curve 2 2.55827 2.31898 Fuel supply curve 3 3.50402 3.17555

[0088] As shown in Table 3, compared with binary search, the search method disclosed in this paper greatly reduces the average number of searches. On a 650MHz ARM Cortex-A7 processor, the time consumption of the four functions CalcuH, CalcuS, CalcuCp, and CalcuK is reduced by more than 39%, while the time consumption of the three functions CalcuS2H, CalcuH2T, and CalcuS2T is not significantly reduced for the same reason as mentioned above.

[0089] As shown in Table 4, when the two methods are applied to the dynamic real-time performance simulation calculation of the turboshaft engine on a 650MHz ARM Cortex-A7 processor, the method disclosed in this patent reduces the time consumption by more than 9.3% compared with the binary search.

[0090] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for finding and interpolating the thermodynamic properties of aircraft engine gas based on memory data, characterized in that: The following steps are involved: 1) Get the input data for the kth search step, including the current target data d k , memory data, where the memory data includes the target data d of the previous search step k-1 and the target data d of the previous search step k-1 Position id in the target data table S k-1 ; 2) Using the recursive search and interpolation method of the thermodynamic properties of the aviation engine gas based on the memory data to search for the target data; the implementation process of step 2) is as follows: 2.1) Get the target data table S: <h2 style=";text-align:left;direction:ltr">S{a1,a2,a3,…,a<h2 style=";text-align:left;direction:ltr"> n <h2 style=";text-align:left;direction:ltr">} Where a n The data is enthalpy, entropy or total temperature. There are n data elements in the table and they are all real numbers. The data are arranged in ascending or descending order. 2.2) Calculate the current target data d k and the target data d of the previous search step k-1 The difference δd k : 2.3) Determine the current target data d k Is it within the target data table S? If not, set the current target data d k Position id in the target data table S k The starting or ending bit of the target data table S: Wherein, 1 is the starting position of the target data table S, and n is the last position of the target data table S; 2.4) Determine the location ID k-1 Whether the location id is satisfied k : In the formula, sort is the position id k-1 Does not satisfy the location id k , then continue searching; 2.5) If the difference δd k ≥0, indicating the current target data d k Target data d of the previous search step k-1 Increasing trend, position id k Must be greater than the location id k-1 , the specific expression is: id k >id k-1 Then use location id k-1 Add 1 as the starting point and search the location ID in ascending order k ; 2.6) If the difference δd k <0, indicating the current target data d k Target data d of the previous search step k-1 Decreasing trend, position id k Must be less than the location id k-1 , the specific expression is: id k ID k-1 Then use location id k-1 Subtract 1 as the starting point, and search for the position ID in descending order k ; 2.7) Save the current target data d found k With location id k , memory data for the next search step.

Images