CN113110558B - Hybrid propulsion unmanned aerial vehicle demand power prediction method - Google Patents
Hybrid propulsion unmanned aerial vehicle demand power prediction method Download PDFInfo
- Publication number
- CN113110558B CN113110558B CN202110517647.1A CN202110517647A CN113110558B CN 113110558 B CN113110558 B CN 113110558B CN 202110517647 A CN202110517647 A CN 202110517647A CN 113110558 B CN113110558 B CN 113110558B
- Authority
- CN
- China
- Prior art keywords
- unmanned aerial
- aerial vehicle
- flight
- state
- altitude
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000003062 neural network model Methods 0.000 claims abstract description 31
- 230000007704 transition Effects 0.000 claims abstract description 28
- 230000009194 climbing Effects 0.000 claims abstract description 26
- 238000012549 training Methods 0.000 claims abstract description 16
- 239000011159 matrix material Substances 0.000 claims abstract description 12
- 210000002569 neuron Anatomy 0.000 claims description 51
- 238000012545 processing Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 230000014509 gene expression Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 230000007423 decrease Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 4
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000011161 development Methods 0.000 abstract description 2
- 230000001737 promoting effect Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 3
- 238000009472 formulation Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/08—Control of attitude, i.e. control of roll, pitch, or yaw
- G05D1/0808—Control of attitude, i.e. control of roll, pitch, or yaw specially adapted for aircraft
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a hybrid propulsion unmanned aerial vehicle demand power prediction method, which comprises the following steps: acquiring flight state information and flight environment information of the unmanned aerial vehicle to generate a first offline data set; establishing an LVQ neural network model capable of identifying working conditions, and training the LVQ neural network model off line; respectively calculating state transition probability matrixes of three state parameters in the stages of takeoff climbing, cruising and descending; judging the current flight phase of the unmanned aerial vehicle, calling a state transition probability matrix corresponding to the flight phase, and predicting the state parameter at the next moment according to three state parameters of the current altitude, flight attack angle and flight speed of the unmanned aerial vehicle; and calculating the flight required power of the unmanned aerial vehicle at the next moment according to the predicted state parameters at the next moment. The method has strong practicability and is beneficial to promoting the development of the energy management strategy technology of the unmanned aerial vehicle based on the demand power prediction.
Description
Technical Field
The invention belongs to the technical field of unmanned aerial vehicles, and particularly relates to a hybrid propulsion unmanned aerial vehicle demand power prediction method.
Background
In recent years, with the gradual maturity of hybrid technologies in the automobile field and the proposal of the concept of 'green aviation' by the country, the hybrid propulsion unmanned aerial vehicle with multiple power sources has gained more and more attention. However, the hybrid propulsion drone technology with multiple power sources is also limited in several ways. In which, multiple factors need to be considered in formulating an energy management strategy for multiple power sources. At present, most unmanned aerial vehicles are based on the overall situation according to flight task profiles when formulating energy management strategies, the formulated strategies ignore the influence of uncertain factors such as variable actual flight conditions, and the single energy management strategy has poor adaptability to hybrid propulsion unmanned aerial vehicles, and is difficult to ensure that each power source works in the best characteristic state to reduce the flight performance.
Because the variability of hybrid propulsion unmanned aerial vehicle flight operating mode, the operating characteristic that consequently the energy management strategy of formulating will each power supply of full play makes its work in high-efficient region to satisfy unmanned aerial vehicle to the demand of dynamic nature. The required power in the flight of the unmanned aerial vehicle is large in change and long in power continuous output time, so that the required power is timely and accurately acquired, and great influence is brought to the formulation of an energy management strategy. The unmanned aerial vehicle can make an energy management strategy in real time according to the predicted required power, and determine the mixing ratio among different power sources, so that the flying stability, economy and safety of the unmanned aerial vehicle can be improved. At present, the prediction method for the power demand of the hybrid propulsion unmanned aerial vehicle is less, and most of the methods are concentrated on hybrid electric vehicles.
Disclosure of Invention
In view of the defects of the prior art, the invention aims to provide a hybrid propulsion unmanned aerial vehicle demand power prediction method to solve the problems that an energy management strategy of the existing unmanned aerial vehicle is difficult to adapt to various flight conditions with uncertainty, and the stability, economy and safety of the unmanned aerial vehicle are poor due to the lack of prediction of demand power. The method provided by the invention can realize the on-line prediction of the required power of the hybrid propulsion unmanned aerial vehicle, the provided LVQ neural network model can judge whether the unmanned aerial vehicle is in a takeoff climbing stage, a cruising stage or a descending stage according to the current flight parameter information of the unmanned aerial vehicle, and call a state transition probability matrix corresponding to the stages according to the clustered flight stages of the unmanned aerial vehicle, and predict the flight state parameter at the next moment according to the flight state parameter of the current unmanned aerial vehicle by using the state transition probability matrix, so as to predict the required power at the next moment of the hybrid propulsion unmanned aerial vehicle, so that theoretical support can be provided for making an energy management strategy to enable multiple power sources of the unmanned aerial vehicle to work under the optimal characteristic, and the safety, the economy and the stability of the flight can be improved.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
the invention discloses a hybrid propulsion unmanned aerial vehicle demand power prediction method, which comprises the following steps:
(1) acquiring flight state information and flight environment information of the unmanned aerial vehicle, generating a first offline data set, and dividing the first offline data set into three offline data sets, namely a data set A, a data set B and a data set C according to takeoff climbing, cruising and descending stages in a task profile;
(2) selecting characteristic parameters capable of expressing the takeoff climbing, cruising and descending stages of the unmanned aerial vehicle, establishing an LVQ neural network model capable of identifying working conditions, selecting the characteristic parameters from a first offline data set, and training the LVQ neural network model offline;
(3) selecting three state parameters of altitude, flight attack angle and flight speed according to the data set A, the data set B and the data set C in the step (1), and respectively calculating state transition probability matrixes of the three state parameters in the stages of takeoff climbing, cruising and descending;
(4) when the unmanned aerial vehicle flies in real time, performing online working condition recognition by using the LVQ neural network model trained in the step (2), judging the current flight stage of the unmanned aerial vehicle, calling a state transition probability matrix corresponding to the flight stage, and predicting the state parameter at the next moment according to the three state parameters of the current altitude, flight angle of attack and flight speed of the unmanned aerial vehicle;
(5) and (4) calculating the flight required power of the unmanned aerial vehicle at the next moment by using the state parameters predicted at the next moment in the step (4).
Further, the flight state information of the unmanned aerial vehicle in the step (1) is the current flight attitude of the unmanned aerial vehicle, and includes: the angle, flight speed and flight angle of attack of the elevator; the flight environment information is the altitude of the position where the unmanned aerial vehicle is located.
Further, the specific steps of the step (1) are as follows:
(11) acquiring flight state information of the unmanned aerial vehicle under a flight task profile by using a sensor carried on the unmanned aerial vehicle, and generating a first offline data set;
(12) selecting the data of the unmanned aerial vehicle in the takeoff and climb stage in the first offline data set in the step (11) to generate a data set A; selecting data of an unmanned aerial vehicle in a cruising stage to generate a data set B; and selecting data of the unmanned aerial vehicle in the descending stage to generate a data set C.
Further, the sensor in the step (11) comprises: GPS, inertial measurement unit, geomagnetic compass, barometric altimeter and electronic gyroscope.
Further, the specific steps of the step (2) are as follows:
(21) selecting characteristic parameters for training the LVQ neural network model, wherein the characteristic parameters comprise an elevator angle and a flight speed of the unmanned aerial vehicle and an altitude of the unmanned aerial vehicle;
(22) respectively carrying out label processing on the selected characteristic parameters, and recording the altitude H position range of the unmanned aerial vehicle as [ H ] in the takeoff and climbing stage of the unmanned aerial vehicle1,H2](ii) a In the cruising stage, the altitude of the unmanned aerial vehicle is small in change, and the range is marked as [ H ]2,H3](ii) a In the descending stage, the altitude range of the unmanned aerial vehicle is recorded as [ H ]3,H4](ii) a In the takeoff and climbing stage of the unmanned aerial vehicle, the angle of the elevator is larger than zero; in the cruising stage, the elevator angle is equal to 0; in the descending stage, the angle of the elevator is less than 0;
wherein H1The initial height of the unmanned aerial vehicle in the takeoff climbing stage is set; h2The lowest height of the unmanned aerial vehicle in the cruising stage is set; h3The highest height of the unmanned aerial vehicle in the cruising stage is set; h4The lowest height of the unmanned aerial vehicle in a descending stage;
(23) according to the processed characteristic parameter data, 75% of the data volume is taken as a training data set, the rest are taken as testing data sets, the sets comprise a plurality of identification samples of different types, and the basic characteristics of three types of flight working conditions can be described;
(24) and constructing an LVQ neural network model, and identifying the current flight stage of the unmanned aerial vehicle according to the current flight state parameters of the unmanned aerial vehicle by the LVQ neural network model.
Further, the step (24) specifically includes:
(241) the LVQ neural network model comprises: an input layer, a competition layer and a linear output layer; selecting m input layer neurons, n competition layer neurons, k output layer neurons, determining the number of the input layer neurons according to the number of parameters recognized in the flight process of the unmanned aerial vehicle, determining the number of the output layer neurons according to the number of working conditions needing to be recognized in the flight process of the unmanned aerial vehicle, and determining the number of the competition layer neurons through tests according to an empirical estimation method; the maximum iteration number is p;
wherein the content of the first and second substances,
(242) the training process of the LVQ neural network model comprises the following steps: initializing a weight ω between an input layer and a competing layerabAnd learning rate eta, the input layer vector is selected as follows:
X=[H,V,θ]T (2)
wherein H is altitude; v is the flight speed; θ is the angle of the elevator;
normalizing the input vector X to obtain a normalized input vector X, sending the normalized input vector X to the input layer, and calculating the distance d between the neuron of the competition layer and the input vectoraThe following were used:
in the formula (d)aDistance between the competition layer neuron and the input vector; omegaabThe weight value between the input layer neuron b and the competition layer neuron a is obtained; x is the number ofbIs an input layer neuron;
selecting the competition layer neuron with the minimum distance with the weight vector as a winning neuron, thereby finding the linear output layer neuron connected with the competition layer neuron; if the corresponding category of the neuron of the linear output layer is consistent with the category of the input vector, the neuron is called correct classification, and the weight is adjusted according to the formula (4); otherwise, called incorrect classification, the weight is adjusted according to the formula (5); entering the next round of training process until the preset maximum iteration number is met;
ωab_new=ωab+η(x-ωab) (4)
ωab_new=ωab-η(x-ωab) (5)
in the formula, ωab_newIs the adjusted weight; eta is the learning rate; and x is the normalized input vector.
Further, the specific steps of the step (3) are as follows:
(31) establishing an unmanned aerial vehicle flight demand power model, wherein a flight resistance calculation formula of the unmanned aerial vehicle in the stages of takeoff climbing, cruising and descending is represented by a formula (6), and the demand power of the unmanned aerial vehicle in flight is represented by a formula (7):
in the formula, S is the wing area of the unmanned aerial vehicle; cDIs a coefficient of resistance; d is the flight resistance of the unmanned aerial vehicle; v∞The flight speed of the unmanned aerial vehicle; prThe required power of the unmanned aerial vehicle during flying is obtained; rho∞The air density of the height of the unmanned aerial vehicle;
(32) determining rho according to the unmanned aerial vehicle flight demand power model established in the step (31)∞、CDAnd V∞The three parameters are important variables influencing the real-time required power change of the unmanned aerial vehicle; separating parameter information of the unmanned aerial vehicle according to the generated offline data sets A, B and C, wherein the selected parameter information comprises altitude H, aircraft attack angle alpha and flying speed V∞Wherein, with doAltitude increase of man-machine place, corresponding altitude density rho∞Decreases, and thus the air density ρ∞Related to altitude H; coefficient of resistance CDRelative to the aircraft angle of attack α;
(33) in the takeoff and climbing phase of the unmanned aerial vehicle, H, alpha and V are determined∞Three state parameter sets are used for state division, and the stage is marked as a stage I;
(34) in the cruising stage of the unmanned aerial vehicle, H, alpha and V are determined∞Three state parameter sets are used for state division, and the stage is marked as stage II;
(35) in the descending stage of the unmanned aerial vehicle, H, alpha and V are determined∞Three state parameter sets are used for state division, and the stage is marked as a stage III;
(36) and (3) processing the actual values of three state parameters of the altitude, the flight angle of attack and the flight speed by using a data preprocessing method so as to be unified with the discretization state set in the steps (33), (34) and (35), wherein the processing is respectively carried out according to the steps (8), (9) and (10):
in the formula, H (t) is the flight height state of the unmanned aerial vehicle at the time t; hrealThe actual flying height at the moment t; alpha (t) is the flight angle of attack state of the unmanned aerial vehicle at the moment t; alpha is alpharealIs the actual flight angle of attack at time t; v (t) is the flight speed state of the unmanned aerial vehicle at the time t; vrealThe actual flying speed at the moment t; ceil () is a ceiling function; floor () is a floor function;
(37) respectively calculating state transition probability matrixes of each state parameter in three flight phases, selecting a statistical state transition probability matrix calculation method, and respectively calculating to obtain state transition probability matrixes of altitude, flight attack angle and flight speed in three flight phases according to an equation (11):
in the formula, PijIs Si→SjThe transition probability of (2); n isijIs Si→SjThe frequency of transfer of (a); n isjTo make a transition from other states to SjTotal frequency of (d); siIs in the current state; sjThe state is the next moment; i and j are any positive integer.
Further, the step (33) specifically includes:
(331) selecting the altitude of the unmanned aerial vehicle to change the altitude according to the step length of 100m to obtain the flying height state set S of the unmanned aerial vehicle1The following were used:
(332) selecting an unmanned aerial vehicle attack angle to divide attack angle states by 0.5-degree step length to obtain a flight attack angle state set S2The following were used:
S2={Si|Si=0.5(i-1),i=1,2,3,…} (13)
(333) selecting flight speed to divide the flight speed state of the unmanned aerial vehicle by 5km/h step length to obtain a flight speed state set S3The following were used:
S3={Si|Si=5(i-1),i=1,2,3,…,25} (14)。
further, the step (34) specifically includes:
(341) selecting an unmanned aerial vehicle to carry out state division by 2m step length to obtain a flying height state set S4The following were used:
(342) selecting an unmanned aerial vehicle attack angle to divide attack angle states by 0.1 degree step length to obtain a flight attack angle state set S5The following were used:
S5={Si|Si=0.1(i-1),i=1,2,3,…} (16)
(343) selecting the flight speed as 2km/h step length to carry out flight speed division to obtain a flight speed state set S6The following were used:
S6={Si|Si=2(i-1)+90,i=1,2,3,…,16} (17)。
further, the step (35) specifically includes:
(351) selecting the altitude of the unmanned aerial vehicle to change the altitude according to the step length of 100m to obtain the flying height state set S of the unmanned aerial vehicle7The following were used:
(352) selecting an unmanned aerial vehicle attack angle to divide attack angle states by 0.5-degree step length to obtain a flight attack angle state set S8The following were used:
S8={Si|Si=0.5(i-1),i=1,2,3,…} (19)
(353) selecting flight speed to divide the flight speed state of the unmanned aerial vehicle by 5km/h step length to obtain a flight speed state set S9The following were used:
S9={Si|Si=5(i-1),i=1,2,3,…,25} (20)。
further, the specific steps of the step (4) are as follows:
(41) when the unmanned aerial vehicle flies in real time, acquiring the angle of an elevator, the flying speed and the altitude data of the unmanned aerial vehicle in the current flying state through a sensor carried on the unmanned aerial vehicle, and carrying out online working condition identification by using a LVQ neural network model to judge the current flying stage of the unmanned aerial vehicle;
(42) and calling a state transition probability matrix corresponding to the flight phase, and selecting a parameter value corresponding to the maximum probability as a state parameter at the next moment according to three acquired state parameters of the current altitude, flight attack angle and flight speed of the unmanned aerial vehicle.
Further, the specific steps of the step (5) are as follows:
according to the predicted altitude H, utilizing H-rho∞Rho is calculated as a relational expression, that is, as expressed by the expressions (21) and (22)∞(ii) a According to the predicted flight incidence angle alpha and the unmanned aerial vehicle CL-alpha characteristic curve, fitting lift coefficient CLThen fitting a resistance coefficient C according to a polar characteristic curve of the unmanned aerial vehicleDCalculating the flight required power of the unmanned aerial vehicle at the next moment according to the formula (7) together with other state parameters;
T=T0-0.0065*H (21)
in the formula, T0The temperature of the air in the standard state is 288.15K; t is the air temperature of the unmanned aerial vehicle with the altitude of H; rho0The air density in the standard state was 1.225kg/m3。
The invention has the beneficial effects that:
1. the method can provide the identification capability of the current working condition of the unmanned aerial vehicle and the prediction capability of the power required at the next moment, and can further improve the stability and the safety of the unmanned aerial vehicle for solving the problem of the influence of uncertain factors on the stability and the safety of the unmanned aerial vehicle in the subsequent flying process of the unmanned aerial vehicle;
2. the method can realize the prediction of the future required power of the unmanned aerial vehicle, thereby providing a power flow distribution basis for the formulation of an energy management strategy of the unmanned aerial vehicle and being beneficial to improving the flight economy of the unmanned aerial vehicle;
3. the method has strong practicability and is beneficial to promoting the development of the energy management strategy technology of the unmanned aerial vehicle based on the demand power prediction.
Drawings
FIG. 1 is a schematic block diagram of the method of the present invention;
FIG. 2 is a schematic structural diagram of an LVQ neural network model;
FIG. 3 shows a UAV CL-a diagram of the characteristic curve;
FIG. 4 is a schematic view of a polar characteristic curve of the UAV;
fig. 5 is a schematic view of a reference coordinate system of the drone.
Fig. 6 is a schematic cross-sectional view of a flight mission of an unmanned aerial vehicle.
Detailed Description
In order to facilitate understanding of those skilled in the art, the present invention will be further described with reference to the following examples and drawings, which are not intended to limit the present invention.
Referring to fig. 1, the hybrid propulsion unmanned aerial vehicle demand power prediction method of the present invention includes the following steps:
(1) acquiring flight state information and flight environment information of the unmanned aerial vehicle, generating a first offline data set, and dividing the first offline data set into three offline data sets, namely a data set A, a data set B and a data set C according to takeoff climbing, cruising and descending stages in a task profile;
the unmanned aerial vehicle flight state information in step (1) is the current flight attitude of the unmanned aerial vehicle, and the method comprises the following steps: the angle, flight speed and flight angle of attack of the elevator; the flight environment information is the altitude of the position where the unmanned aerial vehicle is located;
(11) acquiring flight state information of the unmanned aerial vehicle under a flight task profile by using a sensor carried on the unmanned aerial vehicle, and generating a first offline data set; as shown with reference to FIG. 6;
(12) selecting the data of the unmanned aerial vehicle in the takeoff and climb stage in the first offline data set in the step (11) to generate a data set A; selecting data of an unmanned aerial vehicle in a cruising stage to generate a data set B; and selecting data of the unmanned aerial vehicle in the descending stage to generate a data set C.
Wherein the sensor in step (11) comprises: GPS, inertial measurement unit, geomagnetic compass, barometric altimeter and electronic gyroscope.
(2) Selecting characteristic parameters capable of expressing the takeoff climbing, cruising and descending stages of the unmanned aerial vehicle, establishing an LVQ neural network model capable of identifying working conditions, selecting the characteristic parameters from a first offline data set, and training the LVQ neural network model offline;
the specific steps of the step (2) are as follows:
(21) selecting characteristic parameters for training the LVQ neural network model, wherein the characteristic parameters comprise an elevator angle and a flight speed of the unmanned aerial vehicle and an altitude of the unmanned aerial vehicle;
(22) respectively carrying out label processing on the selected characteristic parameters, and recording the altitude H position range of the unmanned aerial vehicle as [ H ] in the takeoff and climbing stage of the unmanned aerial vehicle1,H2](ii) a In the cruising stage, the altitude of the unmanned aerial vehicle is small in change, and the range is marked as [ H ]2,H3](ii) a In the descending stage, the altitude range of the unmanned aerial vehicle is recorded as [ H ]3,H4](ii) a In the takeoff and climbing stage of the unmanned aerial vehicle, the angle of the elevator is larger than zero; in the cruising stage, the elevator angle is equal to 0; in the descending stage, the angle of the elevator is less than 0;
wherein H1The initial height of the unmanned aerial vehicle in the takeoff climbing stage is set; h2The lowest height of the unmanned aerial vehicle in the cruising stage is set; h3The highest height of the unmanned aerial vehicle in the cruising stage is set; h4The lowest height of the unmanned aerial vehicle in a descending stage;
referring to fig. 5, rotation of the unmanned aerial vehicle around the Y axis indicates that the unmanned aerial vehicle is raising, lowering or flying horizontally, and when data is processed, the angle sign of the elevator corresponding to raising of the unmanned aerial vehicle is marked as positive, the angle sign of the elevator corresponding to lowering of the head is marked as negative, and the angle of the elevator corresponding to flying horizontally is marked as zero;
(23) according to the processed characteristic parameter data, 75% of the data volume is taken as a training data set, the rest are taken as testing data sets, the sets comprise a plurality of identification samples of different types, and the basic characteristics of three types of flight working conditions can be described;
(24) constructing an LVQ neural network model, and identifying the current flight stage of the unmanned aerial vehicle according to the current flight state parameters of the unmanned aerial vehicle by the LVQ neural network model; as shown with reference to FIG. 2;
the step (24) specifically includes:
(241) the LVQ neural network model comprises: an input layer, a competition layer and a linear output layer; selecting m input layer neurons, n competition layer neurons, k output layer neurons, determining the number of the input layer neurons according to the number of parameters recognized in the flight process of the unmanned aerial vehicle, determining the number of the output layer neurons according to the number of working conditions needing to be recognized in the flight process of the unmanned aerial vehicle, and determining the number of the competition layer neurons through tests according to an empirical estimation method; the maximum iteration number is p;
wherein the content of the first and second substances,
(242) the training process of the LVQ neural network model comprises the following steps: initializing a weight ω between an input layer and a competing layerabAnd learning rate eta, the input layer vector is selected as follows:
X=[H,V,θ]T (2)
wherein H is altitude; v is the flight speed; θ is the angle of the elevator;
normalizing the input vector X to obtain a normalized input vector X, sending the normalized input vector X to the input layer, and calculating the distance d between the neuron of the competition layer and the input vectoraThe following were used:
in the formula (d)aDistance between the competition layer neuron and the input vector; omegaabThe weight value between the input layer neuron b and the competition layer neuron a is obtained; x is the number ofbIs an input layer neuron;
selecting the competition layer neuron with the minimum distance with the weight vector as a winning neuron, thereby finding the linear output layer neuron connected with the competition layer neuron; if the corresponding category of the neuron of the linear output layer is consistent with the category of the input vector, the neuron is called correct classification, and the weight is adjusted according to the formula (4); otherwise, called incorrect classification, the weight is adjusted according to the formula (5); entering the next round of training process until the preset maximum iteration number is met;
ωab_new=ωab+η(x-ωab) (4)
ωab_new=ωab-η(x-ωab) (5)
in the formula, ωab_newIs the adjusted weight; eta is the learning rate; and x is the normalized input vector.
(3) Selecting three state parameters of altitude, flight attack angle and flight speed according to the data set A, the data set B and the data set C in the step (1), and respectively calculating state transition probability matrixes of the three state parameters in the stages of takeoff climbing, cruising and descending;
the specific steps of the step (3) are as follows:
(31) establishing an unmanned aerial vehicle flight demand power model, wherein a flight resistance calculation formula of the unmanned aerial vehicle in the stages of takeoff climbing, cruising and descending is represented by a formula (6), and the demand power of the unmanned aerial vehicle in flight is represented by a formula (7):
in the formula, S is the wing area of the unmanned aerial vehicle; cDIs a coefficient of resistance; d is the flight resistance of the unmanned aerial vehicle; v∞The flight speed of the unmanned aerial vehicle; prThe required power of the unmanned aerial vehicle during flying is obtained; rho∞The air density of the height of the unmanned aerial vehicle;
(32) according to the nobody established in the step (31)Determining rho by the model of the power demanded by the aircraft flight∞、CDAnd V∞The three parameters are important variables influencing the real-time required power change of the unmanned aerial vehicle; separating parameter information of the unmanned aerial vehicle according to the generated offline data sets A, B and C, wherein the selected parameter information comprises altitude H, aircraft attack angle alpha and flying speed V∞Wherein, along with the increase of the altitude at which the unmanned aerial vehicle is located, the corresponding altitude density ρ∞Decreases, and thus the air density ρ∞Related to altitude H; coefficient of resistance CDRelative to the aircraft angle of attack α;
(33) in the takeoff and climbing phase of the unmanned aerial vehicle, H, alpha and V are determined∞Three state parameter sets are used for state division, and the stage is marked as a stage I;
(34) in the cruising stage of the unmanned aerial vehicle, H, alpha and V are determined∞Three state parameter sets are used for state division, and the stage is marked as stage II;
(35) in the descending stage of the unmanned aerial vehicle, H, alpha and V are determined∞Three state parameter sets are used for state division, and the stage is marked as a stage III;
(36) and (3) processing the actual values of three state parameters of the altitude, the flight angle of attack and the flight speed by using a data preprocessing method so as to be unified with the discretization state set in the steps (33), (34) and (35), wherein the processing is respectively carried out according to the steps (8), (9) and (10):
in the formula, H (t) is the flight height state of the unmanned aerial vehicle at the time t; hrealFor actual flight at time tA line height; alpha (t) is the flight angle of attack state of the unmanned aerial vehicle at the moment t; alpha is alpharealIs the actual flight angle of attack at time t; v (t) is the flight speed state of the unmanned aerial vehicle at the time t; vrealThe actual flying speed at the moment t; ceil () is a ceiling function; floor () is a floor function;
(37) respectively calculating state transition probability matrixes of each state parameter in three flight phases, selecting a statistical state transition probability matrix calculation method, and respectively calculating to obtain state transition probability matrixes of altitude, flight attack angle and flight speed in three flight phases according to an equation (11):
in the formula, PijIs Si→SjThe transition probability of (2); n isijIs Si→SjThe frequency of transfer of (a); n isjTo make a transition from other states to SjTotal frequency of (d); siIs in the current state; sjThe state is the next moment; i and j are any positive integer.
The step (33) specifically includes:
(331) selecting the altitude of the unmanned aerial vehicle to change the altitude according to the step length of 100m to obtain the flying height state set S of the unmanned aerial vehicle1The following were used:
(332) selecting an unmanned aerial vehicle attack angle to divide attack angle states by 0.5-degree step length to obtain a flight attack angle state set S2The following were used:
S2={Si|Si=0.5(i-1),i=1,2,3,…} (13)
(333) selecting flight speed to divide the flight speed state of the unmanned aerial vehicle by 5km/h step length to obtain a flight speed state set S3The following were used:
S3={Si|Si=5(i-1),i=1,2,3,…,25} (14)。
the step (34) specifically includes:
(341) selecting an unmanned aerial vehicle to carry out state division by 2m step length to obtain a flying height state set S4The following were used:
(342) selecting an unmanned aerial vehicle attack angle to divide attack angle states by 0.1 degree step length to obtain a flight attack angle state set S5The following were used:
S5={Si|Si=0.1(i-1),i=1,2,3,…} (16)
(343) selecting the flight speed as 2km/h step length to carry out flight speed division to obtain a flight speed state set S6The following were used:
S6={Si|Si=2(i-1)+90,i=1,2,3,…,16} (17)。
the step (35) specifically includes:
(351) selecting the altitude of the unmanned aerial vehicle to change the altitude according to the step length of 100m to obtain the flying height state set S of the unmanned aerial vehicle7The following were used:
(352) selecting an unmanned aerial vehicle attack angle to divide attack angle states by 0.5-degree step length to obtain a flight attack angle state set S8The following were used:
S8={Si|Si=0.5(i-1),i=1,2,3,…} (19)
(353) selecting flight speed to divide the flight speed state of the unmanned aerial vehicle by 5km/h step length to obtain a flight speed state set S9The following were used:
S9={Si|Si=5(i-1),i=1,2,3,…,25} (20)。
(4) when the unmanned aerial vehicle flies in real time, performing online working condition recognition by using the LVQ neural network model trained in the step (2), judging the current flight stage of the unmanned aerial vehicle, calling a state transition probability matrix corresponding to the flight stage, and predicting the state parameter at the next moment according to the three state parameters of the current altitude, flight angle of attack and flight speed of the unmanned aerial vehicle;
the specific steps of the step (4) are as follows:
(41) when the unmanned aerial vehicle flies in real time, acquiring the angle of an elevator, the flying speed and the altitude data of the unmanned aerial vehicle in the current flying state through a sensor carried on the unmanned aerial vehicle, and carrying out online working condition identification by using a LVQ neural network model to judge the current flying stage of the unmanned aerial vehicle;
(42) and calling a state transition probability matrix corresponding to the flight phase, and selecting a parameter value corresponding to the maximum probability as a state parameter at the next moment according to three acquired state parameters of the current altitude, flight attack angle and flight speed of the unmanned aerial vehicle.
(5) Calculating the flight required power of the unmanned aerial vehicle at the next moment by using the state parameters predicted at the next moment in the step (4);
the specific steps of the step (5) are as follows:
(51) according to the predicted altitude H, utilizing H-rho∞Rho is calculated as a relational expression, that is, as expressed by the expressions (21) and (22)∞(ii) a According to the predicted flight incidence angle alpha and the unmanned aerial vehicle CL-alpha characteristic curve, fitting lift coefficient CLThen fitting a resistance coefficient C according to a polar characteristic curve of the unmanned aerial vehicleDCalculating the flight required power of the unmanned aerial vehicle at the next moment according to the formula (7) together with other state parameters; as shown with reference to fig. 3 and 4;
T=T0-0.0065*H (21)
in the formula, T0The temperature of the air in the standard state is 288.15K; t is that the altitude of the unmanned aerial vehicle is HThe air temperature of (d); rho0The air density in the standard state was 1.225kg/m3。
While the invention has been described in terms of its preferred embodiments, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.
Claims (5)
1. A hybrid propulsion unmanned aerial vehicle demand power prediction method is characterized by comprising the following steps:
(1) acquiring flight state information and flight environment information of the unmanned aerial vehicle, generating a first offline data set, and dividing the first offline data set into three offline data sets, namely a data set A, a data set B and a data set C according to takeoff climbing, cruising and descending stages in a task profile;
(2) selecting characteristic parameters capable of expressing the takeoff climbing, cruising and descending stages of the unmanned aerial vehicle, establishing an LVQ neural network model capable of identifying working conditions, selecting the characteristic parameters from a first offline data set, and training the LVQ neural network model offline;
(3) selecting three state parameters of altitude, flight attack angle and flight speed according to the data set A, the data set B and the data set C in the step (1), and respectively calculating state transition probability matrixes of the three state parameters in the stages of takeoff climbing, cruising and descending;
(4) when the unmanned aerial vehicle flies in real time, performing online working condition recognition by using the LVQ neural network model trained in the step (2), judging the current flight stage of the unmanned aerial vehicle, calling a state transition probability matrix corresponding to the flight stage, and predicting the state parameter at the next moment according to the three state parameters of the current altitude, flight angle of attack and flight speed of the unmanned aerial vehicle;
(5) calculating the flight required power of the unmanned aerial vehicle at the next moment by using the state parameters predicted at the next moment in the step (4);
the unmanned aerial vehicle flight state information in step (1) is the current flight attitude of the unmanned aerial vehicle, and the method comprises the following steps: the angle, flight speed and flight angle of attack of the elevator; the flight environment information is the altitude of the position where the unmanned aerial vehicle is located;
the specific steps of the step (2) are as follows:
(21) selecting characteristic parameters for training the LVQ neural network model, wherein the characteristic parameters comprise an elevator angle and a flight speed of the unmanned aerial vehicle and an altitude of the unmanned aerial vehicle;
(22) respectively carrying out label processing on the selected characteristic parameters, and recording the altitude H position range of the unmanned aerial vehicle as [ H ] in the takeoff and climbing stage of the unmanned aerial vehicle1,H2](ii) a In the cruising stage, the altitude of the unmanned aerial vehicle is small in change, and the range is marked as [ H ]2,H3](ii) a In the descending stage, the altitude range of the unmanned aerial vehicle is recorded as [ H ]3,H4](ii) a In the takeoff and climbing stage of the unmanned aerial vehicle, the angle of the elevator is larger than zero; in the cruising stage, the elevator angle is equal to 0; in the descending stage, the angle of the elevator is less than 0;
wherein H1The initial height of the unmanned aerial vehicle in the takeoff climbing stage is set; h2The lowest height of the unmanned aerial vehicle in the cruising stage is set; h3The highest height of the unmanned aerial vehicle in the cruising stage is set; h4The lowest height of the unmanned aerial vehicle in a descending stage;
(23) according to the processed characteristic parameter data, 75% of the data volume is taken as a training data set, the rest are taken as testing data sets, the sets comprise a plurality of identification samples of different types, and the basic characteristics of three types of flight working conditions can be described;
(24) constructing an LVQ neural network model, and identifying the current flight stage of the unmanned aerial vehicle according to the current flight state parameters of the unmanned aerial vehicle by the LVQ neural network model;
the specific steps of the step (3) are as follows:
(31) establishing an unmanned aerial vehicle flight demand power model, wherein a flight resistance calculation formula of the unmanned aerial vehicle in the stages of takeoff climbing, cruising and descending is represented by a formula (6), and the demand power of the unmanned aerial vehicle in flight is represented by a formula (7):
in the formula, S is the wing area of the unmanned aerial vehicle; cDIs a coefficient of resistance; d is the flight resistance of the unmanned aerial vehicle; v∞The flight speed of the unmanned aerial vehicle; prThe required power of the unmanned aerial vehicle during flying is obtained; rho∞The air density of the height of the unmanned aerial vehicle;
(32) determining rho according to the unmanned aerial vehicle flight demand power model established in the step (31)∞、CDAnd V∞The three parameters are important variables influencing the real-time required power change of the unmanned aerial vehicle; separating parameter information of the unmanned aerial vehicle according to the generated offline data sets A, B and C, wherein the selected parameter information comprises altitude H, aircraft attack angle alpha and flying speed V∞Wherein, along with the increase of the altitude at which the unmanned aerial vehicle is located, the corresponding altitude density ρ∞Decreases, and thus the air density ρ∞Related to altitude H; coefficient of resistance CDRelative to the aircraft angle of attack α;
(33) in the takeoff and climbing phase of the unmanned aerial vehicle, H, alpha and V are determined∞Three state parameter sets are used for state division, and the stage is marked as a stage I;
(34) in the cruising stage of the unmanned aerial vehicle, H, alpha and V are determined∞Three state parameter sets are used for state division, and the stage is marked as stage II;
(35) in the descending stage of the unmanned aerial vehicle, H, alpha and V are determined∞Three state parameter sets are used for state division, and the stage is marked as a stage III;
(36) and (3) processing actual values of three state parameters of the altitude, the flight attack angle and the flight speed by using a data preprocessing method to be unified with the discretization state set in the steps (33), (34) and (35), wherein the processing is respectively carried out according to the steps (8), (9) and (10):
in the formula, H (t) is the flight height state of the unmanned aerial vehicle at the time t; hrealThe actual flying height at the moment t; alpha (t) is the flight angle of attack state of the unmanned aerial vehicle at the moment t; alpha is alpharealIs the actual flight angle of attack at time t; v (t) is the flight speed state of the unmanned aerial vehicle at the time t; vrealThe actual flying speed at the moment t; ceil () is a ceiling function; floor () is a floor function;
(37) respectively calculating state transition probability matrixes of each state parameter in three flight phases, selecting a statistical state transition probability matrix calculation method, and respectively calculating to obtain state transition probability matrixes of altitude, flight attack angle and flight speed in three flight phases according to an equation (11):
in the formula, PijIs Si→SjThe transition probability of (2); n isijIs Si→SjThe frequency of transfer of (a); n isjTo make a transition from other states to SjTotal frequency of (d); siIs in the current state; sjThe state is the next moment; i and j are any positive integer;
the specific steps of the step (4) are as follows:
(41) when the unmanned aerial vehicle flies in real time, acquiring the angle of an elevator, the flying speed and the altitude data of the unmanned aerial vehicle in the current flying state through a sensor carried on the unmanned aerial vehicle, and carrying out online working condition identification by using a LVQ neural network model to judge the current flying stage of the unmanned aerial vehicle;
(42) calling a state transition probability matrix corresponding to a flight stage, and selecting a parameter value corresponding to the maximum probability as a state parameter at the next moment according to three acquired state parameters of the current altitude, flight attack angle and flight speed of the unmanned aerial vehicle;
the specific steps of the step (5) are as follows:
according to the predicted altitude H, utilizing H-rho∞Rho is calculated as a relational expression, that is, as expressed by the expressions (21) and (22)∞(ii) a According to the predicted flight incidence angle alpha and the unmanned aerial vehicle CL-alpha characteristic curve, fitting lift coefficient CLThen fitting a resistance coefficient C according to a polar characteristic curve of the unmanned aerial vehicleDCalculating the flight required power of the unmanned aerial vehicle at the next moment according to the formula (7) together with other state parameters;
T=T0-0.0065*H (21)
in the formula, T0The temperature of the air in the standard state is 288.15K; t is the air temperature of the unmanned aerial vehicle with the altitude of H; rho0The air density in the standard state was 1.225kg/m3。
2. The hybrid propulsion drone demanded power prediction method according to claim 1, characterized in that said step (24) comprises in particular:
(241) the LVQ neural network model comprises: an input layer, a competition layer and a linear output layer; selecting m input layer neurons, n competition layer neurons, k output layer neurons, determining the number of the input layer neurons according to the number of parameters recognized in the flight process of the unmanned aerial vehicle, determining the number of the output layer neurons according to the number of working conditions needing to be recognized in the flight process of the unmanned aerial vehicle, and determining the number of the competition layer neurons through tests according to an empirical estimation method; the maximum iteration number is p;
wherein the content of the first and second substances,
(242) the training process of the LVQ neural network model comprises the following steps: initializing a weight ω between an input layer and a competing layerabAnd learning rate eta, the input layer vector is selected as follows:
X=[H,V,θ]T (2)
where H is the altitude, V is the flight speed, and θ is the angle of the elevator;
normalizing the input vector X to obtain a normalized input vector X, sending the normalized input vector X to the input layer, and calculating the distance d between the neuron of the competition layer and the input vectoraThe following were used:
in the formula (d)aDistance between the competition layer neuron and the input vector; omegaabThe weight value between the input layer neuron b and the competition layer neuron a is obtained; x is the number ofbIs an input layer neuron;
selecting the competition layer neuron with the minimum distance with the weight vector as a winning neuron, thereby finding the linear output layer neuron connected with the competition layer neuron; if the corresponding category of the neuron of the linear output layer is consistent with the category of the input vector, the neuron is called correct classification, and the weight is adjusted according to the formula (4); otherwise, called incorrect classification, the weight is adjusted according to the formula (5); entering the next round of training process until the preset maximum iteration number is met;
ωab_new=ωab+η(x-ωab) (4)
ωab_new=ωab-η(x-ωab) (5)
in the formula, ωab_newTo be adjustedA weight; eta is the learning rate; and x is the normalized input vector.
3. The hybrid propulsion drone demanded power prediction method according to claim 1, characterized in that said step (33) comprises in particular:
(331) selecting the altitude of the unmanned aerial vehicle to change the altitude according to the step length of 100m to obtain the flying height state set S of the unmanned aerial vehicle1The following were used:
(332) selecting an unmanned aerial vehicle attack angle to divide attack angle states by 0.5-degree step length to obtain a flight attack angle state set S2The following were used:
S2={Si|Si=0.5(i-1),i=1,2,3,…} (13);
(333) selecting flight speed to divide the flight speed state of the unmanned aerial vehicle by 5km/h step length to obtain a flight speed state set S3The following were used:
S3={Si|Si=5(i-1),i=1,2,3,…,25} (14)。
4. the hybrid propulsion drone demanded power prediction method according to claim 1, characterized in that the step (34) comprises in particular:
(341) selecting an unmanned aerial vehicle to carry out state division by 2m step length to obtain a flying height state set S4The following were used:
(342) selecting an unmanned aerial vehicle attack angle to divide attack angle states by 0.1 degree step length to obtain a flight attack angle state set S5The following were used:
S5={Si|Si=0.1(i-1),i=1,2,3,…} (16);
(343) selectingThe flying speed is divided into 2km/h step length to obtain a flying speed state set S6The following were used:
S6={Si|Si=2(i-1)+90,i=1,2,3,…,16} (17)。
5. the hybrid propulsion drone demanded power prediction method according to claim 1, characterized in that said step (35) comprises in particular:
(351) selecting the altitude of the unmanned aerial vehicle to change the altitude according to the step length of 100m to obtain the flying height state set S of the unmanned aerial vehicle7The following were used:
(352) selecting an unmanned aerial vehicle attack angle to divide attack angle states by 0.5-degree step length to obtain a flight attack angle state set S8The following were used:
S8={Si|Si=0.5(i-1),i=1,2,3,…} (19);
(353) selecting flight speed to divide the flight speed state of the unmanned aerial vehicle by 5km/h step length to obtain a flight speed state set S9The following were used:
S9={Si|Si=5(i-1),i=1,2,3,…,25} (20)。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110517647.1A CN113110558B (en) | 2021-05-12 | 2021-05-12 | Hybrid propulsion unmanned aerial vehicle demand power prediction method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110517647.1A CN113110558B (en) | 2021-05-12 | 2021-05-12 | Hybrid propulsion unmanned aerial vehicle demand power prediction method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113110558A CN113110558A (en) | 2021-07-13 |
CN113110558B true CN113110558B (en) | 2022-04-08 |
Family
ID=76722415
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110517647.1A Active CN113110558B (en) | 2021-05-12 | 2021-05-12 | Hybrid propulsion unmanned aerial vehicle demand power prediction method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113110558B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113625741B (en) * | 2021-09-01 | 2022-03-08 | 北京北航天宇长鹰无人机科技有限公司 | Flight segment control method and device |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105701090A (en) * | 2014-10-11 | 2016-06-22 | 中国国际航空股份有限公司 | Simulator for flight management system |
CN106647809A (en) * | 2017-01-06 | 2017-05-10 | 中国航天电子技术研究院 | Energy optimization method for hybrid power system of UAV based on task planning |
CN106970523A (en) * | 2017-03-01 | 2017-07-21 | 南京航空航天大学 | A kind of energy management strategies of aircraft self-adapting power and heat management system |
CN109917810A (en) * | 2019-04-12 | 2019-06-21 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle cluster energy management method |
CN112758335A (en) * | 2021-01-20 | 2021-05-07 | 南京航空航天大学 | Series-parallel hybrid power unmanned aerial vehicle power system and control method thereof |
Family Cites Families (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6672415B1 (en) * | 1999-05-26 | 2004-01-06 | Toyota Jidosha Kabushiki Kaisha | Moving object with fuel cells incorporated therein and method of controlling the same |
JP2010287028A (en) * | 2009-06-11 | 2010-12-24 | Sony Corp | Information processor, information processing method and program |
FR3012230B1 (en) * | 2013-10-17 | 2016-01-01 | Airbus Operations Sas | METHOD AND APPARATUS FOR REVIEWING A LOW ALTITUDE FLIGHT PHASE OF A FLIGHT TRACK OF AN AIRCRAFT |
US10586464B2 (en) * | 2015-07-29 | 2020-03-10 | Warren F. LeBlanc | Unmanned aerial vehicles |
CN106768123A (en) * | 2016-11-30 | 2017-05-31 | 中国直升机设计研究所 | A kind of depopulated helicopter fuel oil predictor method |
CN106892124B (en) * | 2017-01-23 | 2018-12-07 | 北京瑞深航空科技有限公司 | Hybrid power unmanned plane |
CN108898696B (en) * | 2018-07-03 | 2021-01-26 | 中国民航大学 | Aircraft power system security analysis method based on dynamic fault tree and Markov |
CN111061281B (en) * | 2018-10-16 | 2023-08-15 | 海鹰航空通用装备有限责任公司 | Aircraft flight scheme generation method and generation system and aircraft with same |
CN109240323B (en) * | 2018-11-02 | 2021-07-13 | 北京控制工程研究所 | Aerospace vehicle reentry guidance method capable of analyzing structure in real time |
CN110046735B (en) * | 2018-12-10 | 2021-02-26 | 南京航空航天大学 | Aircraft departure fuel oil consumption evaluation method based on flight data analysis |
CN111625013B (en) * | 2019-02-27 | 2023-04-25 | 杭州海康机器人技术有限公司 | Unmanned aerial vehicle flight task execution energy prediction method and device |
CN109871034A (en) * | 2019-03-25 | 2019-06-11 | 苏州极目机器人科技有限公司 | Flight control method, device and unmanned vehicle |
CN110348595B (en) * | 2019-05-31 | 2022-12-23 | 南京航空航天大学 | Unmanned aerial vehicle hybrid propulsion system energy management and control method based on flight data |
CN110276479B (en) * | 2019-05-31 | 2023-01-03 | 南京航空航天大学 | Cruise phase fuel consumption prediction method for aircraft mass change |
CN110610260B (en) * | 2019-08-21 | 2023-04-18 | 南京航空航天大学 | Driving energy consumption prediction system, method, storage medium and equipment |
CN111461292B (en) * | 2020-03-17 | 2023-06-27 | 南京航空航天大学 | Unmanned aerial vehicle real-time track prediction method |
CN112257151A (en) * | 2020-10-20 | 2021-01-22 | 南京航空航天大学 | Aircraft flight staging identification system |
CN112671451B (en) * | 2020-12-10 | 2022-11-18 | 北京邮电大学 | Unmanned aerial vehicle data collection method and device, electronic device and storage medium |
-
2021
- 2021-05-12 CN CN202110517647.1A patent/CN113110558B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105701090A (en) * | 2014-10-11 | 2016-06-22 | 中国国际航空股份有限公司 | Simulator for flight management system |
CN106647809A (en) * | 2017-01-06 | 2017-05-10 | 中国航天电子技术研究院 | Energy optimization method for hybrid power system of UAV based on task planning |
CN106970523A (en) * | 2017-03-01 | 2017-07-21 | 南京航空航天大学 | A kind of energy management strategies of aircraft self-adapting power and heat management system |
CN109917810A (en) * | 2019-04-12 | 2019-06-21 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle cluster energy management method |
CN112758335A (en) * | 2021-01-20 | 2021-05-07 | 南京航空航天大学 | Series-parallel hybrid power unmanned aerial vehicle power system and control method thereof |
Also Published As
Publication number | Publication date |
---|---|
CN113110558A (en) | 2021-07-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110348595B (en) | Unmanned aerial vehicle hybrid propulsion system energy management and control method based on flight data | |
CN108107911B (en) | Solar airplane autonomous optimization flight path planning method | |
CN107944648B (en) | Large ship speed and oil consumption rate prediction method | |
CN111008504B (en) | Wind power prediction error modeling method based on meteorological pattern recognition | |
CN104156594B (en) | Dynamic flight station-crossing time estimation method based on Bayes network | |
CN109191922A (en) | A kind of extensive four-dimensional dynamic track prediction method and device | |
CN111680875B (en) | Unmanned aerial vehicle state risk fuzzy comprehensive evaluation method based on probability baseline model | |
CN111461292B (en) | Unmanned aerial vehicle real-time track prediction method | |
CN106768123A (en) | A kind of depopulated helicopter fuel oil predictor method | |
CN111176263B (en) | Online identification method for thrust fault of aircraft based on BP neural network | |
CN113777931B (en) | Icing wing type pneumatic model construction method, device, equipment and medium | |
CN113110558B (en) | Hybrid propulsion unmanned aerial vehicle demand power prediction method | |
CN109144099B (en) | Fast evaluation method for unmanned aerial vehicle group action scheme based on convolutional neural network | |
US11142337B2 (en) | Method and system for determining a descent profile | |
CN107016880A (en) | A kind of continuously declining into nearly oil consumption optimization method based on simulated annealing | |
CN112506218A (en) | Reentry aircraft any no-fly-off area fly-around method based on intelligent trajectory prediction | |
CN114610070A (en) | Unmanned aerial vehicle-cooperated wind power plant intelligent inspection method | |
CN115204041A (en) | Large aircraft approach and approach stage oil consumption prediction method | |
CN107368090A (en) | A kind of fixed-wing solar energy unmanned plane endurance method of estimation | |
CN117037006B (en) | Unmanned aerial vehicle tracking method with high endurance capacity | |
CN105118332B (en) | A kind of air traffic control analog simulation method for detecting abnormality and device based on clustering methodology | |
CN110414407B (en) | Ship propeller air entrainment identification method based on evidence reasoning rule and self-adaptive lifting | |
Guan et al. | A study of 4D trajectory prediction based on machine deep learning | |
CN117113870A (en) | Aerodynamic modeling method of fixed wing aircraft considering wind interference | |
CN112925344B (en) | Unmanned aerial vehicle flight condition prediction method based on data driving and machine learning |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |