Aircraft platform ambient temperature prediction method and system based on dynamic optimization
By partitioning the aircraft and combining the Kalman dynamic filter recursive equation, the coefficient matrix of the temperature prediction model is dynamically adjusted, which solves the problems of large error and lag in the temperature prediction of aircraft platform environment by traditional heat transfer analysis models, and realizes high-precision temperature prediction.
Patent Information
- Application Number
- CN202311628442.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-11-30
- Publication Date
- 2026-02-06
- Estimated Expiration
- 2043-11-30
AI Technical Summary
Existing traditional heat transfer analysis models have limited generalization ability in predicting ambient temperature on aircraft platforms, resulting in large prediction errors and lag, and making it impossible to dynamically adjust model parameters.
A dynamic optimization-based approach is adopted to divide the aircraft into a third zone, establish a heat flow equation and solve it discretizedly, and combine it with the Kalman dynamic filter recursive equation to dynamically adjust the coefficient matrix of the temperature prediction model and improve the prediction accuracy.
It effectively improves the accuracy of aircraft platform environmental temperature prediction, eliminates prediction lag, has good model extrapolation stability, and has small prediction error.
Smart Images
Figure CN118194426B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of environmental engineering, in particular to an aircraft platform environmental temperature prediction method and system based on dynamic optimization. BACKGROUND
[0002] Temperature is the main reason affecting equipment performance and inducing failure, and high temperature, low temperature and rapid temperature change will have various adverse effects on equipment, and temperature is also a "multiplier" of other environmental effects, for example, the corrosion rate of metal will accelerate in high temperature environment. For aircraft, aerodynamic heating in flight, radiation heating of the aircraft surface by the sun and self-heating of internal electronic equipment make the temperature environment of the aircraft platform more severe and the influence more serious.
[0003] The most direct method to determine the platform environmental temperature at present is field environmental measurement, but for extreme states that cannot be obtained by measurement, a temperature prediction model needs to be established based on measured data and combined with heat transfer analysis method to determine the full-process temperature value. However, the traditional heat transfer analysis model is mostly based on regression method to determine a fixed set of model coefficients, and the model generalization ability is limited, resulting in large prediction error and prediction result lag, and the model parameters cannot be dynamically predicted. SUMMARY
[0004] The purpose of the present application is to provide an aircraft platform environmental temperature prediction method and system based on dynamic optimization, which effectively improves the prediction accuracy and eliminates the prediction lag.
[0005] An aircraft platform environmental temperature prediction method based on dynamic optimization, comprising:
[0006] S1, partitioning the aircraft based on physical structure, heat transfer characteristics and measurement point data:
[0007] S11, dividing the cabins with the same or similar physical structure in the aircraft into the same region to obtain a plurality of first partitions;
[0008] S12, dividing the first partitions with the same or similar heat transfer characteristics into the same region to obtain a plurality of second partitions;
[0009] S13, dividing the second partitions with the same or similar measurement point data into the same region to obtain a plurality of third partitions;
[0010] S2, establishing a heat flow equation for each third partition, discretizing each heat flow equation, and solving based on actual data to obtain a temperature prediction equation for each third partition;
[0011] S3, establishing an ambient temperature prediction model of each of the third partitions, the ambient temperature prediction model being as follows:
[0012]
[0013] wherein i=[1, 2, Λ, N], N is the total number of the third partitions, is a predicted ambient temperature of the i-th third partition at time t, is a state coefficient of the i-th third partition at time t, is a state coefficient of the i-th third partition at time t-1, is a random disturbance of the i-th third partition from time t-1 to time t, is a difference matrix between the predicted temperature of the i-th third partition at time t and the recovery temperature and the predicted temperatures of the remaining third partitions, the predicted temperature being obtained based on a temperature prediction equation, e i is a measurement error of the i-th third partition at time t;
[0014] S4, solving each of the ambient temperature prediction models based on a Kalman dynamic filtering recursive equation to obtain a predicted ambient temperature of each of the third partitions.
[0015] Optionally, the Kalman dynamic filtering recursive equation is as follows:
[0016]
[0017]
[0018]
[0019]
[0020]
[0021]
[0022]
[0023] wherein T is a transpose, e i is a prediction error of the i-th third partition, is an actual platform ambient temperature of the i-th third partition at time t, is an error variance matrix of, W i is a variance of the random disturbance ω, R i is a sum of the white noise, σ i is an error variance matrix of the platform ambient temperature value of the i-th third partition, A ian intermediate quantity for the i-th third partition, an error variance matrix, an inverse matrix of σ i i a variance of e i a difference matrix between the temperature prediction value of the i-th third partition at the t+1 moment and the recovery temperature and the temperature prediction values of the remaining third partitions, a predicted ambient temperature of the i-th third partition at the t+1 moment.
[0024] Optionally, the heat flow equation is:
[0025]
[0026] wherein S is the number of regions exchanging heat with the i-th third partition, Q i is the heat increment of the i-th third partition, Q j is the heat exchanged between the j-th region and the i-th third partition.
[0027] The application further provides an aircraft platform ambient temperature prediction system based on dynamic optimization, which comprises:
[0028] a partition module for partitioning an aircraft based on physical structures, heat transfer characteristics and measurement point data; the partition module comprises:
[0029] a first partition unit for dividing cabins with the same or similar physical structures in the aircraft into the same region to obtain a plurality of first partitions;
[0030] a second partition unit for dividing the first partitions with the same or similar heat transfer characteristics into the same region to obtain a plurality of second partitions;
[0031] a third partition unit for dividing the second partitions with the same or similar measurement point data into the same region to obtain a plurality of third partitions;
[0032] a partition temperature prediction module for establishing heat flow equations of the third partitions, discretizing the heat flow equations and solving the equations based on actual data to obtain temperature prediction equations of the third partitions;
[0033] an ambient temperature prediction model module for establishing ambient temperature prediction models of the third partitions; the ambient temperature prediction model is as follows:
[0034]
[0035] wherein i=[1, 2, Λ, N], N is the total number of third partitions, is a predicted value of the environmental temperature of the i-th third partition at time t, is a state coefficient of the i-th third partition at time t, is a state coefficient of the i-th third partition at time t-1, is a random disturbance of the i-th third partition from time t-1 to time t, is a difference matrix between the predicted value of the temperature of the i-th third partition at time t and the recovered temperature and the predicted values of the temperatures of the remaining third partitions, the predicted value of the temperature being obtained based on a temperature prediction equation, is a measurement error of the i-th third partition at time t;
[0036] a platform environmental temperature prediction module configured to solve each of the environmental temperature prediction models based on a Kalman dynamic filtering recursive equation to obtain the predicted values of the environmental temperatures of the third partitions.
[0037] Optionally, the Kalman dynamic filtering recursive equation is:
[0038]
[0039]
[0040]
[0041]
[0042]
[0043]
[0044]
[0045] wherein T is a transpose, e i is a prediction error of the i-th third partition, is an actual value of the platform environmental temperature of the i-th third partition at time t, is an error variance matrix of, W i is a variance of the random disturbance ω, R i is a sum of and white noise, σ i is an error variance matrix of the platform environmental temperature value of the i-th third partition, A i is an intermediate quantity in calculation of the i-th third partition, is an error variance matrix of, V is an inverse matrix of σ i , V i is a variance of e i , is a difference matrix between the temperature prediction value of the i th third partition at t+1 time and the recovery temperature and the temperature prediction value of the remaining third partitions, is the environmental temperature prediction value of the i th third partition at t+1 time.
[0046] Optionally, the heat flow equation is:
[0047]
[0048] In the formula, S is the number of regions exchanging heat with the i th third partition, Q i is the heat increment of the i th third partition, Q j is the heat exchanged between the j th region and the i th third partition.
[0049] Effects of the present application are as follows:
[0050] The aircraft platform environmental temperature prediction method based on dynamic optimization of the present application solves the model generalization problem by using the dynamic temperature prediction model based on Kalman dynamic filtering, and converts the single fixed coefficient of the heat transfer model into a dynamic coefficient matrix, so that the model extrapolation stability is good.
[0051] The aircraft platform environmental temperature prediction method based on dynamic optimization of the present application fits the measured data, effectively improves the model prediction accuracy, and eliminates the prediction hysteresis. BRIEF DESCRIPTION OF DRAWINGS
[0052] Figure 1 is a flow chart of the aircraft platform environmental temperature prediction method based on dynamic optimization of the present application;
[0053] Figure 2 is a comparison chart of the environmental temperature prediction value and the actual value of the first cabin of an example unmanned aerial vehicle platform;
[0054] Figure 3 is an error diagram of the environmental temperature prediction value of the first cabin of an example unmanned aerial vehicle platform;
[0055] Figure 4 is a comparison chart of the error of the environmental temperature prediction value of the method of the present application and the conventional heat transfer analysis model;
[0056] Figure 5 is an error distribution chart of the environmental temperature prediction value of the method of the present application and the conventional heat transfer analysis model. DETAILED DESCRIPTION
[0057] Hereinafter, embodiments of the present application will be described with reference to the accompanying drawings.
[0058] Figure 1is a flow chart of an aircraft platform ambient temperature prediction method based on dynamic optimization according to the present application. As shown in Figure 1 The present application provides an aircraft platform ambient temperature prediction method based on dynamic optimization, which comprises:
[0059] S1, partitioning the aircraft based on physical structure, heat transfer characteristics and measurement point data:
[0060] S11, dividing the cabins with the same or similar physical structure in the aircraft into the same region to obtain a plurality of first partitions. That is, the cabins with the same or similar physical structure are merged into one first partition.
[0061] S12, dividing the first partitions with the same or similar heat transfer characteristics into the same region to obtain a plurality of second partitions. That is, the first partitions with the same or similar heat transfer characteristics are merged into one second partition.
[0062] S13, dividing the second partitions with the same or similar measurement point data into the same region to obtain a plurality of third partitions. That is, the second partitions with the same or similar measurement point data are merged into one third partition.
[0063] S2, establishing the heat flow equation of each third partition, discretizing each heat flow equation, and solving based on actual data to obtain the temperature prediction equation of each third partition.
[0064] The heat flow equation is:
[0065]
[0066] In the formula, S is the number of regions exchanging heat with the i-th third partition, Q i is the heat increment of the i-th third partition, Q j is the heat exchanged between the j-th region and the i-th third partition.
[0067] S3, establishing the ambient temperature prediction model of each third partition. The ambient temperature prediction model is as follows:
[0068]
[0069] In the formula, i=[1, 2, Λ, N], N is the total number of third partitions, is the ambient temperature prediction value of the i-th third partition at t, is the state coefficient of the i-th third partition at t, is the state coefficient of the i-th third partition at t-1, is the random disturbance of the i-th third partition from t-1 to t, is a difference matrix between the temperature prediction value of the i th third partition at t moment and the recovery temperature and the temperature prediction value of the rest third partitions, the temperature prediction value is obtained based on a temperature prediction equation, e i is a measurement error of the i th third partition at t moment;
[0070] S4, each environment temperature prediction model is solved based on a Kalman dynamic filtering recursive equation, and the environment temperature prediction value of each third partition is obtained.
[0071] The Kalman dynamic filtering recursive equation is:
[0072]
[0073]
[0074]
[0075]
[0076]
[0077]
[0078]
[0079] Wherein, T is transposition, e i is a prediction error of the i th third partition, is an actual value of the platform environment temperature of the i th third partition at t moment, is Error variance matrix of, W i is the variance of random disturbance ω, R i is Sum of white noise, σ i is the error variance matrix of the platform environment temperature value of the i th third partition, A i is a calculation intermediate quantity of the i th third partition, is the error variance matrix of , is the inverse matrix of σ i , V i is the variance of e i , is a difference matrix between the temperature prediction value of the i th third partition at t+1 moment and the recovery temperature and the temperature prediction value of the rest third partitions, is an environment temperature prediction value of the i th third partition at t+1 moment.
[0080] The application also provides an aircraft platform environment temperature prediction system based on dynamic optimization, which comprises:
[0081] a partition module configured to partition the aircraft based on the physical structure, heat transfer characteristics and measurement point data.
[0082] The partition module comprises:
[0083] a first partition unit configured to divide cabin rooms with the same or similar physical structure in the aircraft into the same region, to obtain a plurality of first partitions.
[0084] a second partition unit configured to divide the first partitions with the same or similar heat transfer characteristics into the same region, to obtain a plurality of second partitions.
[0085] a third partition unit configured to divide the second partitions with the same or similar measurement point data into the same region, to obtain a plurality of third partitions.
[0086] a partition temperature prediction module configured to establish heat flow equations for the third partitions, discretize the heat flow equations, and solve the heat flow equations based on actual data to obtain temperature prediction equations for the third partitions.
[0087] an ambient temperature prediction model module configured to establish ambient temperature prediction models for the third partitions. The ambient temperature prediction model is as follows:
[0088]
[0089] wherein i = [1, 2, Λ, N], N is the total number of third partitions, is the ambient temperature prediction value of the i-th third partition at time t, is the state coefficient of the i-th third partition at time t, is the state coefficient of the i-th third partition at time t-1, is the random disturbance of the i-th third partition from time t-1 to time t, is the difference matrix between the temperature prediction value of the i-th third partition at time t and the recovery temperature and the temperature prediction values of the remaining third partitions, the temperature prediction value being obtained based on the temperature prediction equation, e i is the measurement error of the i-th third partition at time t.
[0090] a platform ambient temperature prediction module configured to solve the ambient temperature prediction models based on a Kalman dynamic filtering recursive equation to obtain the ambient temperature prediction values of the third partitions.
[0091] The Kalman dynamic filtering recursive equation is as follows:
[0092]
[0093]
[0094]
[0095]
[0096]
[0097]
[0098]
[0099] wherein T is transpose, e i is the prediction error of the i th third partition, is the actual value of the platform ambient temperature of the i th third partition at t time, is the error variance matrix of , W i is the variance of random disturbance ω, R i is the error variance matrix of and white noise, σ i is the error variance matrix of the platform ambient temperature value of the i th third partition, A i is the calculation intermediate quantity of the i th third partition, is the error variance matrix of , is the inverse matrix of σ i , V i is the variance of e i , is the difference matrix between the temperature prediction value of the i th third partition at t+1 time and the recovery temperature and the temperature prediction value of the remaining third partitions, is the ambient temperature prediction value of the i th third partition at t+1 time.
[0100] The heat flow equation is:
[0101]
[0102] wherein S is the number of regions exchanging heat with the i th third partition, Q i is the heat increment of the i th third partition, Q j is the heat exchanged between the j th region and the i th third partition.
[0103] Specifically, the method of the present application is described below with a certain unmanned aerial vehicle platform (UAV) as an example.
[0104] Based on the physical structure, heat transfer characteristics and measurement point data, the unmanned aerial vehicle platform is divided into three third partitions, which are defined as the first cabin, the second cabin and the third cabin respectively.
[0105] Firstly, the temperature environment in the cabin is analyzed. For the first cabin, the main heat transfer processes include the heat transfer between the boundary layer airflow and the air in the first cabin, the heat generated by electronic equipment, the energy and mass transport caused by the gas exchange between the first cabin and the outside, the air flow between the cabins, and the structural heat conduction between adjacent cabins. Since most of the first cabin is on the lower surface of the UAV platform, the solar radiation heat transfer is omitted. According to the analysis results, the heat flow equation of the first cabin is as follows:
[0106]
[0107] wherein C A is the heat capacity of the first cabin, is the heat transfer amount between the boundary layer airflow and the first cabin at time t, Q m is the heat transfer amount of the gas exchange between the first cabin and the outside at time t, Q AB is the heat transfer amount between the first cabin and the second cabin at time t, Q AC is the heat transfer amount between the first cabin and the third cabin at time t, Q A is the heat generated by the equipment in the first cabin at time t, T A is the temperature value of the first cabin at time t.
[0108]
[0109] wherein R OA is the heat transfer coefficient between the external environment and the first cabin, T r is the temperature of the boundary layer at time t.
[0110] Q m (t)=m 1A (t)c a T r (t)-m 2A (t)c a T r (t);
[0111] wherein m 1A is the air mass flow rate entering the first cabin from the outside at time t, m 2A is the air mass flow rate leaking from the first cabin to the outside at time t, c a is the specific heat capacity of air.
[0112] Q AB (t)=R AB (T B (t)-T A (t))+C a m AB (t)(T B(t) = T A (t) ;
[0113] wherein R AB is the heat transfer coefficient between the first chamber and the second chamber, T B (t) is the temperature value of the second chamber at time t, m AB (t) is the mass exchange between the first chamber and the second chamber at time t.
[0114] Q AC (t) = R AC (T C (t) - T A (t)) + C a m AC (t) (T C (t) - T A (t)) ;
[0115] wherein R AC is the heat transfer coefficient between the first chamber and the third chamber, T C (t) is the temperature value of the third chamber at time t, m AC (t) is the mass exchange between the first chamber and the third chamber at time t.
[0116] Since no airflow control is adopted in the UAV, the air exchange between the chambers can be regarded as natural convection, so it can be assumed that the air exchange mass flow rate between the first chamber, the second chamber and the third chamber remains unchanged, which is a constant value. At the same time, it is assumed that Q A (t) remains unchanged, then the heat flow equation of the first chamber is:
[0117]
[0118] wherein R 1A = R 0A + m AB c a , R 2A = R AB + m AB c a , R 3A = R AC + m AC c a , R 4A = Q A , Q A is the heat dissipation power of the electronic device.
[0119] The heat flow equations of the second chamber and the third chamber are the same as those of the first chamber.
[0120] After discretization, we get:
[0121] T A(t+1)-T A (t) = C1(T r (t)-T A (t))+C2(T B (t)-T A (t))+C3(T C (t)-T A (t))+C4;
[0122] C1 = (AtR 1A ) / C a , C2 = (AtR 2A ) / C a , C3 = (AtR 3A ) / C a , C4 = (AtR 4A ) / C a , At is time step,
[0123] The actual data is introduced, and the parameters are obtained by using the multiple linear regression method, as shown in Table 1.
[0124] Table 1 Parameter table
[0125] Parameter [C1] [C2] [C3] [C4] Value 0.00094 0.00033 -0.000085 -0.00075
[0126] As can be seen from Table 1, the value of C3 is very small, indicating that the heat exchange between the first cabin and the third cabin is small.
[0127] Since T r (t) = (1 + 0.178M(t) 2 )(T0(t) + 273) - 273, T0(t) = T e (t) - 0.0065(H(t) + H0), the temperature prediction equation of the first cabin can be obtained as:
[0128]
[0129] In the formula: M(t) is the flight Mach number of the unmanned aerial vehicle platform at time t, T0(t) is the atmospheric temperature at the height at time t, T e (t) is the local atmospheric temperature at time t, H(t) is the flight altitude at time t, and H0 is the local altitude of the flight.
[0130] Based on the above formula, the temperature prediction value of the first cabin at t+1 time can be obtained, and the same for the second cabin and the third cabin. Based on the difference between the temperature prediction value of the first cabin and the recovery temperature, the difference between the temperature prediction value of the first cabin and the temperature prediction value of the third cabin, and the difference between the temperature prediction value of the first cabin and the temperature prediction value of the second cabin, the following is constructed
[0131] establishing an environment temperature prediction model of each of the third partitions.
[0132] Solving each of the environment temperature prediction models based on Kalman dynamic filtering recursive equations to obtain the environment temperature prediction value of the first cabin, the environment temperature prediction value of the second cabin and the environment temperature prediction value of the third cabin.
[0133] Since b is a state coefficient, the initial value b0 of b is (-0.00075, 0.00094, 0.00033, -0.000085), C0 is the covariance matrix of b0, and b0 is assumed to be completely accurate, so C0 is a zero matrix. To determine the W matrix, the regression coefficient matrix b is recalculated by using the temperature difference of different flights t-1 = (0.00681, 0.00102, 0.00029, -0.00049), and the W matrix is determined by using the formula: The W matrix is determined by the change of b t-1 , b0.
[0134]
[0135] p and q are the values of the rows and columns in the W matrix.
[0136] V is the residual sum of squares of the regression equation, and V = 10. In this way, the initial values of the Kalman dynamic filtering recursive equations are all determined, and the environment temperature prediction value can be obtained based on the Kalman dynamic filtering recursive equations.
[0137] The environment temperature prediction value of the first cabin and the actual value of the platform environment temperature are as shown in Figure 2 , and the prediction error is as shown in Figure 3 From Figure 3 , it can be seen that the maximum error is only 2.76℃, the prediction curve has good followability, and there is no change lag phenomenon.
[0138] The error comparison chart of the environment temperature prediction value of the method of the present application and the traditional heat transfer analysis model is as shown in Figure 4 , the error distribution chart of the environment temperature prediction value of the method of the present application and the traditional heat transfer analysis model is as shown in Figure 5 The maximum error of the present application is 3.19℃, and the average error is 0.99℃. Assuming that the error obeys the normal distribution of [-0.56, 1.02], it is calculated that the maximum prediction error under the confidence probability of 95% is 2.6℃, the maximum prediction error of the traditional heat transfer analysis model is 6.78℃, and obviously the present application has better effect.
[0139] The above embodiments are only used to describe the preferred embodiments of the present application, and are not used to limit the scope of the present application. Without departing from the design spirit of the present application, various modifications and improvements made by those skilled in the art to the technical solutions of the present application shall fall within the protection scope of the present application as defined by the claims.
Claims
1. A dynamic optimization based aircraft platform ambient temperature prediction method, characterized by, It comprises: S1, partitioning the aircraft based on physical structure, heat transfer characteristics and measurement point data: S11, dividing the cabins with the same or similar physical structure in the aircraft into the same area to obtain a plurality of first partitions; S12, dividing the first partitions with the same or similar heat transfer characteristics into the same area to obtain a plurality of second partitions; S13, dividing the second partitions with the same or similar measurement point data into the same area to obtain a plurality of third partitions; S2, establishing a heat flow equation for each third partition, discretizing each heat flow equation, and solving based on actual data to obtain a temperature prediction equation for each third partition; S3, establishing an ambient temperature prediction model for each third partition, the ambient temperature prediction model being as follows: ； In the formula: N is the total number of third partitions, is the predicted value of the environmental temperature of the ith third partition at time t, is the state coefficient of the ith third partition at time t, , is the state coefficient of the ith third partition at time t-1, is the random disturbance of the ith third partition from time t-1 to time t, is the difference matrix between the temperature predicted value of the ith third partition at time t and the recovery temperature and the temperature predicted values of the remaining third partitions, the temperature predicted value being obtained based on a temperature prediction equation, is the measurement error of the ith third partition at time t; S4, solving each ambient temperature prediction model based on a Kalman dynamic filtering recursive equation to obtain an ambient temperature prediction value for each third partition.
2. The dynamic optimization based aircraft platform ambient temperature prediction method of claim 1, wherein, The Kalman dynamic filtering recursive equation is: ； ； ； ； ； ； ； where T is transpose, is the prediction error of the i-th third partition, is the actual value of the platform ambient temperature of the i-th third partition at time t, is the error variance matrix of , is the variance of the random disturbance , is the sum of and white noise, is the error variance matrix of the platform ambient temperature value of the i-th third partition, is the calculation intermediate quantity of the i-th third partition, is the error variance matrix of , is the inverse matrix of , is the variance of , is the difference value matrix between the temperature prediction value of the i-th third partition at time t+1 and the recovery temperature and the temperature prediction values of the remaining third partitions, is the ambient temperature prediction value of the i-th third partition at time t+1.
3. The dynamic optimization based aircraft platform ambient temperature prediction method of claim 1, wherein, The heat flow equation is: ； where S is the number of regions exchanging heat with the ith third partition, is the heat increment of the ith third partition, is the heat exchanged between the jth region and the ith third partition.
4. A dynamic optimization based aircraft platform ambient temperature prediction system, characterized by, It comprises: A partition module for partitioning the aircraft based on physical structure, heat transfer characteristics and measurement point data; The partition module comprises: A first partition unit for dividing the cabins with the same or similar physical structure in the aircraft into the same area to obtain a plurality of first partitions; A second partition unit for dividing the first partitions with the same or similar heat transfer characteristics into the same area to obtain a plurality of second partitions; A third partition unit for dividing the second partitions with the same or similar measurement point data into the same area to obtain a plurality of third partitions; A partition temperature prediction module for establishing a heat flow equation for each third partition, discretizing each heat flow equation, and solving based on actual data to obtain a temperature prediction equation for each third partition; An ambient temperature prediction model module for establishing an ambient temperature prediction model for each third partition; the ambient temperature prediction model being as follows: ； In the formula: N is the total number of third partitions, is the predicted value of the ambient temperature of the ith third partition at time t, is the state coefficient of the ith third partition at time t, , is the state coefficient of the ith third partition at time t-1, is the random disturbance of the ith third partition from time t-1 to time t, is the difference matrix between the predicted value of the ith third partition at time t and the recovery temperature and the predicted values of the other third partitions, the predicted value being obtained based on a temperature prediction equation, is the measurement error of the ith third partition at time t; A platform ambient temperature prediction module for solving each ambient temperature prediction model based on a Kalman dynamic filtering recursive equation to obtain an ambient temperature prediction value for each third partition.
5. The dynamic optimization based aircraft platform ambient temperature prediction system of claim 4, wherein, The Kalman dynamic filtering recursive equation is: ； ； ； ； ； ； ； In the formula, T is the transpose. Let be the prediction error for the i-th third partition. Let be the actual value of the platform environment temperature in the i-th third partition at time t. for The error variance matrix, For random perturbation variance for The sum of white noise, It is the error variance matrix of the platform ambient temperature value of the i-th third partition. For the calculation of the i-th third partition, for The error variance matrix, for The inverse matrix, for variance Let be the difference matrix between the predicted temperature of the i-th third partition at time t+1 and the recovered temperature, as well as the predicted temperature of the remaining third partitions. Let t+1 be the predicted environmental temperature value of the i-th third partition.
6. The dynamic optimization based aircraft platform ambient temperature prediction system of claim 4, wherein, The heat flow equation is: ； where S is the number of regions exchanging heat with the ith third partition, is the heat increment for the ith third partition, is the heat exchanged between the jth region and the ith third partition.