Helicopter pilot control intention fuzzy inference sway angle feedback anti-sway method

Abstract

The present application provides a helicopter pilot manipulation intention fuzzy reasoning sling swing angle feedback anti-swing method, comprising: step 1, establishing a helicopter sling flight dynamics model; step 2, establishing a flight control system model integrating a sling swing angle feedback control law CAF and a fuzzy module, coupling and integrating the helicopter sling flight dynamics model to construct a helicopter sling flight closed-loop system model; step 3, detailed design of the fuzzy module, obtaining pilot manipulation intention by using a fuzzy reasoning method according to the sling flight state and pilot manipulation history; step 4, real-time adjusting the sling swing angle feedback control law configuration according to the pilot manipulation intention to eliminate conflicts, and verifying the effectiveness through human-machine-sling coupled closed-loop system simulation. The present application has good command tracking effect and good sling swing damping, and can improve the manned helicopter sling flight quality.

Description

TECHNICAL FIELD

[0001] The present application relates to a method for reducing the sway of a helicopter pilot's control intention fuzzy reasoning sling angle feedback. BACKGROUND

[0002] Current helicopter slung load flight damping techniques can be divided into two categories, the first category directly increases the stability of the slung load, for example, the CONEX container slung load is equipped with additional tail wing to improve its stability (reference: Raz R, Rosen A, Carmeli A, Lusardi J, Cicolani LS, Robinson D. Wind tunnel and flight evaluation of passive stabilization of a cargo container slung load. Journal of the American Helicopter Society, 2010; 55(3):32001-3200118).The second type of method takes the motion control of the hoisting point as the core, and the technical means adopted include: 1) installing special mechanical devices (such as active cargo hooks, reference: Enciu J, Singh A, Horn JF. Stabilization of external loads in high-speed flight using an active cargo hook. Journal of the American Helicopter Society, 2020; 65(2): 1-12) on the belly of the helicopter; 2) integrating hoisting swing control technology (such as cable angle feedback, reference: Ivler C, Tischler M, Powell JD. Cable angle feedback control systems to improve handling qualities for helicopters with slung loads. In: AIAA Guidance, Navigation, and Control Conference, AIAA Guidance, Navigation, and Control Conference. Portland, Oregon: American Institute of Aeronautics and Astronautics. 2011) in the flight control feedback loop; 3) integrating trajectory generation technology (such as input shaping, reference: Adams C, Potter J, Singhose W. Input-shaping and model-following control of a helicopter carrying a suspended load. Journal of Guidance, Control, and Dynamics, 2015; 38(1): 94-105) in the flight control forward path.In recent years, to further solve the problem of poor slung load flight quality, additional task tailoring based on cable angle feedback (Ref: Ivler CM, Powell JD, Tischler MB, Fletcher JW, Ott C. Design and flight test of a cable angle feedback flight control system for the RASCAL JUH-60 helicopter. Journal of the American Helicopter Society, 2014; 59(4): 1-15), rod quantity judgment (Ref: Nonnenmacher D, Kim HM. Evaluation of an advanced slung load control system for piloted cargo operations. CEAS Aeronautical Journal, 2020; 11(4): 897-915) or combination with active hook control (Ref: Patterson BW, Enns DR, King C. Design and flight test of a hybrid external load stabilization system for an H-6 helicopter testbed. In: Proceedings of the American Helicopter Society 71st Annual Forum. Virginia Beach, Virginia: American Helicopter Society. 2015) technical methods have emerged.

[0003] The first type of helicopter slung load flight oscillation reduction technology principle is simple, direct and effective, only needs to study the isolated slung load, but needs to carry out a large number of wind tunnel tests, has poor universality, high cost, and is only suitable for the unified optimization of large quantities of slung loads. In the second type of slung point motion oscillation reduction technology, installing special mechanical devices on the belly of the helicopter increases the complexity and cost of the system, and the effect is limited by the mechanical device; the flight control feedback loop integrated oscillation reduction reduces the command tracking performance; the flight control forward path integrated oscillation reduction avoids exciting the slung load oscillation by making the helicopter and the slung point move in a specific way, but the motion trajectory of the human-controlled helicopter is mainly determined by the pilot. This method may conflict with the pilot's control intention. Therefore, the existing slung point motion oscillation reduction technology is difficult to coordinate the requirements of slung load oscillation reduction effect and command tracking performance, and a method is needed to coordinate the oscillation reduction, but the existing technology is still difficult to effectively coordinate the multiple requirements. SUMMARY

[0004] The present application aims to solve the technical problem that the sling damping effect and the command tracking performance conflict when a person hangs on a helicopter, and under the background of the prior art, the two are difficult to coordinate, and can only be considered by trade-off, resulting in degradation of flight quality when the helicopter is flying, and additional burden on the pilot. In order to solve this problem, the present application uses a fuzzy reasoning method to judge the pilot's manipulation intention in real time, and adjusts the sling swing angle feedback configuration in real time, so that the sling flight response characteristics are consistent with the pilot's manipulation intention, and the conflict between the sling damping effect and the command tracking performance is eliminated. The present application specifically provides a helicopter pilot manipulation intention fuzzy reasoning sling swing angle feedback damping method, comprising the following steps:

[0005] Step 1, a helicopter sling flight dynamics model is established;

[0006] Step 2, a flight control system model integrating the sling swing angle feedback control law CAF and the fuzzy module is established, coupled and integrated with the helicopter sling flight dynamics model, to build a helicopter sling flight closed-loop system model;

[0007] Step 3, the fuzzy module is designed in detail, and the pilot's manipulation intention is obtained using a fuzzy reasoning method according to the sling flight state and the pilot's manipulation history;

[0008] Step 4, the pilot's manipulation intention is used to adjust the sling swing angle feedback control law configuration in real time to eliminate the conflict, and the effectiveness is verified through human-machine-sling coupled closed-loop system simulation.

[0009] In step 1, the helicopter sling flight dynamics model includes a rotor model, a helicopter rigid body dynamics model, an aerodynamic interference model, a sling model and a sling object rigid body dynamics model.

[0010] In step 1, the inputs of the helicopter sling flight dynamics model are the rotor automatic tilting deflection and the tail rotor pitch, and the outputs are the helicopter rigid body motion state and the sling swing angle derived from the sling system.

[0011] In step 1, the rotor blades are set as rigid bodies, the inertia load, the aerodynamic load and the damping moment of the blades are added to the balance at the hinge, and the flapping motion equation and the pitching motion equation of the i-th blade are established, as shown in the following formula:

[0012]

[0013]

[0014] wherein β i , denote the i-th blade flap angle, flap angular velocity and flap angular acceleration, respectively; denote the i-th blade pitch angle, pitch angular velocity and pitch angular acceleration, respectively; e is the hinge offset; b and I b denote the blade static moment and the inertia moment, respectively; Ω, denote the rotor speed and the rotational acceleration, respectively; is the three-axis acceleration of the hub; p s , q s , r s is the three-axis angular velocity of the hub, is the three-axis angular acceleration of the hub; and are the flap and the pitch aerodynamic moments, respectively; and are the pitch damper's flap and pitch moments of force.

[0015] In Step 1, the state-space form of the Helicopter and Slung Load System (HSLS) model is:

[0016]

[0017] where x and are the state vector of the HSLS model and its derivative, respectively; f is the state equation; δ is the control input vector, and t is time, as follows:

[0018]

[0019] δ = (δ col , δ lat , δ lon , δ tr ) T

[0020] where v0, v 1C , v 1S are the rotor dynamic inflow state quantities; v 0TR is the tail rotor uniform inflow state quantity; u, v, w are the three-axis velocity vectors of the helicopter; p, q, r are the three-axis angular velocities of the helicopter; φ, θ, ψ are the three attitude angles of the helicopter; x, y, z are the three-axis positions of the helicopter in the ground axis system; u L , v L , w L are the three-axis velocity vectors of the slung load; p L , q L , r L are the three-axis angular velocities of the slung load; φ L , θ L, ψ L are three attitude angles of the sling; x L , y L , z L are three position coordinates of the sling in the body axis system; δ col , δ lat , δ lon , δ tr denote the collective, lateral cyclic, longitudinal cyclic and tail rotor collective of the helicopter respectively; T denotes the matrix transpose.

[0021] In Step 1, the sling swing angle is determined according to the relative motion between the helicopter and the sling, and the longitudinal swing angle α E and the lateral swing angle β K are defined in the sling hook coordinate system (O-XYZ) c which is consistent with the ground axis system (O-XYZ) c , the origin O K is the average position of the sling point on the helicopter, and the relative position of the origin O K in the body axis (O-XYZ) B is defined as denotes the position of the center of gravity of the sling in the sling hook coordinate system; in order to obtain the swing angle, the sling hook coordinate system is expressed by the swing angle and the HSLS state quantity respectively, and the following relationship is established:

[0022]

[0023] wherein, A KB is the conversion matrix from the body axis to the sling hook coordinate system; l c is the distance from the average sling point on the helicopter to the center of gravity of the sling.

[0024] In Step 2, the transfer function G of the classical explicit model tracking MFCS control system is:

[0025]

[0026] wherein, M denotes the command model of the classical explicit model tracking MFCS control system, P -1 denotes the inverse model of the classical explicit model tracking MFCS control system, H denotes the feedback loop of the classical explicit model tracking MFCS control system; Pτ is the controlled object, wherein τ is the equivalent time delay of high-order dynamics.

[0027] In Step 2, under the requirement of ADS-33E-PRF, the ideal helicopter flight dynamics and flight quality are obtained by designing the command model, the roll channel and the pitch channel adopt the attitude command and attitude holding response type, and the vertical channel and the yaw channel adopt the speed command and height or direction holding response type.​

[0028] Adding the sling swing angle feedback loop L and fuzzy rule Z to the classical explicit model following MFCS control system;

[0029] Coupling the helicopter sling system model, the classical explicit model following MFCS control system, the sling swing angle feedback loop and the fuzzy rule as the helicopter sling flight closed loop system model.

[0030] In step 3, the structure of the man-machine cooperative fuzzy sling swing angle feedback is established, and the fuzzy module is designed in detail. The sling longitudinal swing angle α c and angular velocity are fed back to the longitudinal channel, the lateral swing angle β c and angular velocity are fed back to the lateral channel; the sling swing angle needs to be processed by a high-pass filter to trim the steady-state swing angle, and the fuzzy module receives the lateral swing angle β c , angular velocity and pilot lateral control δ lat,p , and obtains the time-varying lateral swing angle rate gain K Specifically, the following steps are included:

[0031] Step 3-1, input preprocessing: based on the undamped pendulum model, an equivalent swing angle algorithm is proposed, as shown in the following formula, β c and are processed as the equivalent lateral swing angle β c,e , that is, the maximum swing angle corresponding to the swing rate of zero:

[0032]

[0033] Where m L is the mass of the sling, and g is the acceleration of gravity;

[0034] Then use the moving average algorithm for β c,e , as shown in the following formula, to obtain the average equivalent lateral swing angle within a few seconds as the first input variable of the fuzzy reasoning system, at the same time, the absolute value δ lat,p of the pilot lateral control input is taken as the second input variable, and the moving average value of the absolute value of the lateral control input is taken as the third input variable

[0035] Where g represents the gravity; T β and T lat represent the integral time of β c,e and the integral time of |δ lat,p | respectively;

[0036] Step 3-2, fuzzification: Fuzzification involves three fuzzy inputs: average equivalent lateral swing angle |δ lat,p | and its moving average Divide swing angle into: VS very small, S small, M medium, L large, VL very large, divide pilot input into S small, M medium and L large according to the magnitude of pilot input, divide pilot control intention into three categories: make the sling swing stabilize as soon as possible STB, make the helicopter track command input CMD and balance the sling swing damping and command tracking performance BLC;

[0037] Step 3-3, fuzzy rule:

[0038] Based on average equivalent lateral swing angle |δ lat,p | and its moving average Apply fuzzy rule to infer pilot control intention;

[0039] Step 3-4, defuzzification:

[0040] Determine variable by applying fuzzy rule Membership degree under existing input value, get

[0041]

[0042] Where, N, S i ,v i , respectively represent fuzzy grade, membership degree and fuzzy value;

[0043] Step 3-5, output post-processing: add gain rising damper after fuzzy output, for first order inertia link, to realize Slow increase, if Reduce, directly provided to the controller without passing through damping link.

[0044] In step 4, the pilot, control system and helicopter sling system constitute the whole man-machine-sling coupled closed loop system;

[0045] Pilot model adopts Hess's structured pilot model, including proprioceptive feedback loop, vestibular feedback loop and visual feedback loop;

[0046] Control element is represented by Y C , corresponding to closed loop control system composed of helicopter sling flight dynamics model and flight control system, Y NM represents pilot's neuromuscular system, Y FS represents helicopter's force sensing system. Y PFFor the pilot body sensation characteristic, based on the rudder deflection δ F Get the body sensation control Y VF For the pilot vestibular system characteristic, used to sense acceleration or angular velocity, and output the vestibular control U VF ; Y P Indicate the outer loop compensation of the pilot, based on the position error to determine the desired attitude C, and then through the inner loop compensation Y E And the central system time delay τ0, get the pilot visual compensation control U based on the visual error of the attitude E ;

[0047] After establishing the man-machine-hangar coupled closed loop system, the task subject element specified in the helicopter flight quality specification ADS-33E is used for numerical simulation, the effectiveness of the real-time configuration of the sling swing angle feedback is judged according to the simulation results, and the parameters are adjusted according to the response results.

[0048] The application also provides a storage medium storing a computer program or instructions, which realizes the helicopter pilot manipulation intention fuzzy reasoning sling swing angle feedback sway reduction method when the computer program or instructions are run.

[0049] Beneficial effects: the prior art is difficult to coordinate the command tracking performance and the sling sway reduction effect of the helicopter sling flight, and can only trade off between the two, even if there is a method trying to dynamically switch the control law corresponding to the command tracking performance and the sling sway reduction effect, but it may not conform to the pilot's manipulation intention, and the present application uses the fuzzy reasoning method to infer the pilot's manipulation intention in real time, and uses it as the basis for adjusting the sway control law, which has better man-machine cooperation than the existing sway reduction method, and improves the helicopter sling flight quality. BRIEF DESCRIPTION OF DRAWINGS

[0050] The above and / or other aspects of the present application will become more apparent by describing in detail the present application with reference to the accompanying drawings, as follows:

[0051] Figure 1 It is a technical solution flowchart.

[0052] Figure 2 It is a schematic diagram of a nonlinear helicopter sling flight dynamics model.

[0053] Figure 3 It is a schematic diagram of the sling swing angle definition.

[0054] Figure 4 It is a schematic diagram of the UH-60A+CONEX preflight trim characteristic.

[0055] Figure 5 It is a schematic diagram of the UH-60A helicopter preflight trim characteristic.

[0056] Figure 6a is the frequency response characteristic diagram of UH-60A+CONEX.

[0057] Figure 6b is the frequency response characteristic diagram of UH-60A+CONEX.

[0058] Figure 7 is the explicit model tracking control diagram of the man-machine collaborative fuzzy sling angle feedback.

[0059] Figure 8 is the man-machine collaborative fuzzy sling angle feedback structure diagram.

[0060] Figure 9 is the internal structure diagram of the fuzzy module.

[0061] Figure 10a is the membership function diagram.

[0062] Figure 10b is the membership function diagram.

[0063] Figure 10c is the membership function diagram.

[0064] Figure 11 is the man-machine-sling coupling closed-loop system model diagram.

[0065] Figure 12 is the structured pilot model diagram.

[0066] Figure 13a is the helicopter roll response in the lateral positioning subject.

[0067] Figure 13b is the helicopter roll response in the lateral positioning subject. DETAILED DESCRIPTION

[0068] To solve the technical problem that the sling anti-sway effect and the command tracking performance conflict when a manned helicopter is in sling flight, the helicopter pilot manipulation intention fuzzy reasoning sling angle feedback anti-sway method is provided, as shown in Figure 1 As shown in the figure, the overall technical scheme of the present application is shown.

[0069] Firstly, the helicopter sling flight dynamics model is established and verified, as shown in Figure 2 As shown in the figure, the helicopter sling flight dynamics model structure is shown, taking the UH-60A helicopter and the CONEX container as an example, the helicopter sling flight closed-loop system model is constructed, and there are 44 states, and the state space form is as follows:

[0070]

[0071] where x and are the state vector and its derivative of the HSLS model respectively; f is the state equation; δ is the control input vector, t is time, as follows:

[0072]

[0073] δ = (δ col ,δ lat ,δ lon ,δ tr ) T

[0074] where v0, v 1C , v 1S are the rotor dynamic inflow state variables; v 0TR is the tail rotor uniform inflow state variable; u, v, w are the helicopter's three-axis velocity vectors; p, q, r are the helicopter's three-axis angular velocities; φ, θ, ψ are the helicopter's three attitude angles; x, y, z are the helicopter's three-axis positions in the ground axis system; u L , v L , w L are the sling's three-axis velocity vectors; p L , q L , r L are the sling's three-axis angular velocities; φ L , θ L , ψ L are the sling's three attitude angles; x L , y L , z L are the sling's three-axis positions in the ground axis system; δ col , δ lat , δ lon , δ tr represent the helicopter's collective, lateral cyclic, longitudinal cyclic and tail rotor collective respectively; T represents matrix transpose. After the model is established, the model correctness needs to be verified, the present application uses the flight test data of UH-60A helicopter and UH-60A helicopter with CONEX container slings to verify, the verification results are as follows Figure 4 is the UH-60A helicopter with CONEX container's forward flight trim characteristics, Figure 5 is the UH-60A helicopter forward flight trim characteristics, Figure 6a is the UH-60A helicopter with CONEX container's lateral frequency response characteristics in hovering, Figure 6b is the lateral frequency response characteristics at 30 knots forward flight, the results show that the model calculation results are basically consistent with the flight test results.

[0075] Secondly, the flight control system model of the integrated cable angle feedback control law (CAF) and fuzzy logic is established, coupled with the helicopter sling flight dynamics model to build the helicopter sling flight closed-loop system model. As shown in Figure 7 Fig. 1 is a structure of the flight control system, which is based on the classical model-following control system (MFCS) widely used by modern helicopters. The classical MFCS is composed of a command model M, an inverse model P -1 and a feedback loop H. The helicopter sling system is the controlled object Pτ, where τ is the equivalent time delay of high-order dynamics, including rotor dynamics, actuator dynamics, etc., δ and x are the control quantity and state quantity, respectively, x cmd is the command input state quantity, δ inv is the output of the inverse model, Δx is the state error, Δδ L is the control increment of the cable angle feedback loop, and Δδ is the control increment of the body feedback loop. The present application adds the cable angle feedback loop L and the fuzzy rule Z to the classical MFCS.

[0076] Then, the pilot's manipulation intention is obtained by using the fuzzy inference method according to the sling flight state and the pilot's manipulation history. For the pilot's manipulation intention recognition, the fuzzy inference system is established based on the common sense consistent with human intuition. The pilot's manipulation intention can be divided into three categories: making the sling swing stable as soon as possible (STB), making the helicopter track the command input (CMD) and balancing the sling swing damping and the command tracking performance (BLC). As shown in Figure 9 Fig. 2, the pilot's manipulation intention is judged by input preprocessing, fuzzy inference system and output post-processing according to the current manipulation amplitude, the average manipulation amplitude in the past period of time and the current sling swing degree. Figure 9 In the formula, β c and are the cable lateral swing angle and the lateral swing angular velocity, respectively, is the average equivalent lateral swing angle in seconds, is the moving average value, δ lat,p is the pilot's lateral manipulation input, is the moving average value of the absolute value of the lateral manipulation input, is the lateral swing angular velocity gain.

[0077] Finally, the cable angle feedback control law configuration is adjusted in real time according to the pilot's manipulation intention to eliminate the conflict, and the effectiveness is verified by the simulation of the man-machine-sling coupled closed-loop system. For the real-time configuration of the cable angle feedback parameters, the existing research results show that only by adjusting the lateral swing angular velocity gain The performance of the instruction tracking or the damping of the swing of the sling can be close to the optimum, and since the longitudinal and lateral swings are coupled when the sling swings, the longitudinal swing is also improved when the lateral swing is suppressed. Therefore, the pilot's manipulation intention is adjusted in real time based on fuzzy inference The parameters of the sling swing angle feedback are configured in real time. As shown in Figure 11 The structure of the man-machine-sling coupled closed loop system is shown, numerical simulation is carried out by using the task subject element specified in the helicopter flight quality specification ADS-33E, and the effectiveness of the real-time configuration of the sling swing angle feedback is judged according to the simulation results.

[0078] Embodiment

[0079] The embodiment provides a sling swing angle feedback swing reduction method based on fuzzy inference of helicopter pilot's manipulation intention, which comprises the following steps:

[0080] (1) Helicopter flight dynamics modeling

[0081] Figure 2 The model in the formula (1) is composed of a rotor model, a helicopter rigid body dynamics model, an aerodynamic interference model, a sling model and a sling body dynamics model. The inputs of the model are the rotor automatic tilting deflection and the tail rotor total pitch, and the outputs are the helicopter rigid body motion state and the sling swing angle derived from the sling system.

[0082] The model considers the rigid body dynamics, the rotor high-order dynamics and the aerodynamic interference. The induced velocity is based on a three-dimensional dynamic inflow model. The nonlinear aerodynamic load of the rotor blade is based on the blade element theory. On the basis of the wind tunnel test data, the interpolation method is used to obtain the airfoil lift coefficient and the drag coefficient of the blade element. The invention sets the rotor blade as a rigid body, and the inertia load, the aerodynamic load and the damping moment of the blade are added to the balance at the hinge to establish the flapping motion equation and the edgewise motion equation of the i-th blade, as shown in the following formula:

[0083]

[0084]

[0085] In the formula, β i , is the flapping angle of the i-th blade, the flapping angle velocity, and the flapping angle acceleration; is the edgewise angle of the i-th blade, the edgewise angle velocity, and the edgewise angle acceleration; e is the hinge offset amount. S b and I b are the blade static moment and the inertia moment; Ω, is the rotor speed and the rotation acceleration; is the hub line acceleration; p s , q s , r s , is the hub angular velocity and angular acceleration; and are the flap and edgewise aerodynamic moments, respectively; and are the edgewise damper effects on flap and edgewise moments, respectively.

[0086] The aerodynamic interference models consider the rotor downwash on the fuselage, tail rotor, horizontal tail and vertical tail, the fuselage wake effects on the tail and the aerodynamic interference between the tails. These aerodynamic interference models can be represented as additional inflow in the wind tunnel test data. The tail rotor aerodynamic model is based on the rotor disc theory, and the other components aerodynamic models are based on the nonlinear wind tunnel test data interpolation. The sling load is connected to the helicopter sling point by linear elastic damping slings, and the sling tension can be calculated according to the relative motion of the sling point in different sling configurations. The sling load aerodynamic model is based on the nonlinear wind tunnel test data interpolation, and the additional inflow is calculated based on the rotor downwash test data.

[0087] This embodiment takes the UH-60A helicopter and CONEX container as the research object. The main rotor of the UH-60A has 4 blades, so the HSLS model has 44 states, and the state space form is as follows:

[0088]

[0089] where x and are the state vector and its derivative of the HSLS model, respectively; f is the state equation; δ is the control input vector, and t is the time, as follows:

[0090]

[0091] δ = (δ col , δ lat , δ lon , δ tr ) T

[0092] where v0, v 1C , v 1S are the rotor dynamic inflow state quantities; v 0TR is the tail rotor uniform inflow state quantity; u, v, w are the three-axis velocity vectors of the helicopter; p, q, r are the three-axis angular velocities of the helicopter; φ, θ, ψ are the three attitude angles of the helicopter; x, y, z are the three-axis positions of the helicopter in the ground axis system; u L , v L , w L are the three-axis velocity vectors of the sling load; p L , q L , r L are the three-axis angular velocities of the sling load; φ L , θ L,ψ L These are the three attitude angles of the suspended object; x L ,y L ,z L It refers to the three-axis position of the suspended object under the ground axis system; δ col ,δ lat ,δ lon ,δ tr These represent the collective pitch, lateral pitch, longitudinal pitch, and tail rotor collective pitch of the helicopter, respectively; T represents matrix transpose.

[0093] The sling angle is determined by the relative motion between the helicopter and the suspended load. For example... Figure 3 The diagram shown illustrates the definition of the sling swing angle, relative to the earth axis (O-XYZ). E A consistent hook coordinate system (O-XYZ) K The longitudinal swing angle α is defined in the text. c and lateral swing angle β c . Origin O K This is the average position of the suspension point on the helicopter. Origin O K On the body axis (O-XYZ) B The relative position in is defined as This indicates the position of the center of gravity of the suspended object in the hook coordinate system. To obtain the swing angle, [the following is used:] The following equation represents the relationship between the pendulum angle and HSLS state variables:

[0094]

[0095] In the formula, A KB It is the transformation matrix from the body axis to the hook coordinate system; l c It is the average distance from the sling point on the helicopter to the center of gravity of the suspended object.

[0096] This invention uses flight test data from a UH-60A helicopter and a UH-60A helicopter with CONEX container hoists to verify the accuracy of the model. The verification results are as follows: Figure 4 , Figure 5 , Figure 6a , Figure 6b As shown, the results indicate that the model calculation results are in good agreement with the flight test results.

[0097] (2) Modeling of helicopter sling-mounted flight closed-loop system

[0098] The helicopter sling flight control system model established in this invention integrates the Cable Angle Feedback (CAF) control law and fuzzy module, and is coupled and integrated with the helicopter sling flight dynamics model to construct a helicopter sling flight closed-loop system model.

[0099] As Figure 7 The figure shows the structure of flight control system, which is based on the classic Model-Following Control System (MFCS) widely used in modern helicopters. The classic MFCS consists of a command model M, an inverse model P -1 and a feedback loop H. The helicopter sling system is the controlled object Pτ, where τ is the equivalent time delay of high-order dynamics, including rotor dynamics, actuator dynamics, etc.

[0100] A big advantage of MFCS is that the bandwidth of the system is determined by the command model, and the transfer function is as follows: if there is an ideal inverse model P -1 , the dynamic characteristics of the controlled object can be completely eliminated, and the response characteristics of the closed-loop system are completely determined by the command model M and the equivalent time delay τ. However, it is not realistic to completely eliminate high-frequency dynamic characteristics, so the inverse model is usually fitted with a first-order relationship element.

[0101]

[0102] Under the requirements of ADS-33E-PRF, the ideal helicopter flight dynamics and flight quality can be obtained by designing the command model. The roll channel and pitch channel of the embodiment adopt the attitude command and attitude hold response type (ACAH), and the vertical channel and yaw channel adopt the rate command and height / direction hold response type (RCHH / RCDH).

[0103] The present application adds a sling swing angle feedback loop L and a fuzzy module Z to the classic MFCS. The current CAF and its improved methods are mainly for helicopters and secondarily for pilots. Pilots need to adapt to the CAF switching logic set by humans, which not only brings additional training costs, but more importantly, may violate the pilot's control intuition. In fact, CAF essentially suppresses sling swing by moving the helicopter, and if the pilot tries to keep the helicopter position unchanged at this time, it will cause a conflict between the pilot's control intention and CAF. The human-machine cooperative fuzzy CAF proposed in this embodiment is mainly for pilots and secondarily for helicopters, which allows CAF to identify the pilot's current control intention and configure internal parameters in real time, optimizes the command tracking performance when the pilot wants to control the helicopter, and optimizes the sling swing damping when the pilot wants to suppress the sling swing. To this end, the pilot's control intention needs to be identified, and the CAF parameters need to be configured to achieve optimal command tracking performance or optimal sling swing damping.

[0104] (3) Pilot control intention fuzzy reasoning and sling swing angle feedback real-time adjustment method

[0105] For pilot control intention recognition, the present application establishes a fuzzy inference system based on common sense consistent with human intuition, and judges the control intention according to the current control amplitude of the pilot, the average control amplitude in the past period of time and the current hanging swing degree. For example, if the current hanging swing is relatively mild, and the current control of the pilot is more rapid than in the past period of time, it indicates that the pilot hopes to have better command tracking performance; if the current hanging swing is relatively severe, and the current control of the pilot is more gentle than in the past period of time, it indicates that the pilot hopes to have better hanging swing damping.

[0106] For real-time adjustment of CAF parameters, the existing research results show that only by adjusting the lateral swing angle rate gain , the command tracking performance or the hanging swing damping can be close to optimal. Since the longitudinal and lateral swings are coupled when the hanging swing occurs, the longitudinal swing is also improved when the lateral swing is suppressed. Therefore, the pilot control intention inferred by the fuzzy inference system is used to adjust the CAF parameters in real time in the present embodiment. Real-time configuration of CAF parameters and man-machine cooperation are realized.

[0107] Figure 8 The structure of the man-machine cooperative fuzzy CAF is composed of a CAF and a fuzzy module. The hanging longitudinal swing angle α c and angular velocity are fed back to the longitudinal channel, and the lateral swing angle β c and angular velocity are fed back to the lateral channel. The hanging swing angle needs to be processed by a high-pass filter to trim the steady-state swing angle. The fuzzy module receives the lateral swing angle β c and angular velocity and the pilot lateral control δ lat,p , and obtains the time-varying lateral swing angle rate gain

[0108] As shown in Figure 9 , the internal structure of the fuzzy module is used to adjust the lateral swing angle gain in real time. The inputs of the module include three main parameters: the lateral swing angle β c and angular velocity and the pilot lateral control δ lat,p . These inputs need to be preprocessed before being transmitted to the fuzzy inference system, and the pilot control intention is inferred in the fuzzy inference system. Since the lateral swing angle rate gain output by the fuzzy inference system is rapidly changing, the gain change needs to be post-processed to make the gain change more consistent with human intuition

[0109] Input preprocessing:

[0110] The input needs to be pre-processed before it can be transmitted to the fuzzy inference system. To avoid the periodicity of the sway angle, the equivalent sway angle algorithm is proposed based on the undamped pendulum model, as shown in the following equation, which is c and is processed as the equivalent lateral sway angle c,e , which is the maximum sway angle corresponding to the sway angle rate of zero.

[0111]

[0112] Subsequently, the c,e is obtained using the moving average algorithm, as shown in the following equation, which is the average equivalent lateral sway angle in several seconds as the first input variable of the fuzzy inference system. At the same time, the absolute value of the pilot's lateral control input lat,p is taken as the second input variable, and the moving average value of the absolute value of the lateral control input

[0113]

[0114] where g represents the gravity; T β and T lat are the integral times of c,e and |δ lat,p | respectively.

[0115] Fuzzification:

[0116] Fuzzification refers to the conversion of numerical values or generalized parameters into fuzzy sets for reasoning through fuzzy logic. In this embodiment, fuzzification involves three fuzzy inputs: the average equivalent lateral sway angle the absolute value of the pilot's lateral control |δ lat,p | and its moving average value For the system configuration of the UH-60A+CONEX, it is generally considered that a sway angle exceeding 30 to 40 degrees is dangerous, so the sway angle is divided into: VS (very small), S (small), M (medium), L (large), and VL (very large), with the membership function as shown in Figure 10a For the pilot's input, it is divided into S (small), M (medium), and L (large) according to the magnitude of the pilot's input, with the membership function as shown in Figure 10b The pilot's control intention can be divided into three categories: stabilizing the swing of the sling as soon as possible (STB), making the helicopter track the command input (CMD), and balancing the swing of the sling and the command tracking performance (BLC). They are closely related to the lateral sway angle rate gain For the UH-60A+CONEX, A value greater than 3 is sufficient to stabilize the suspension swing (STB); a value less than 0.5 optimizes command tracking performance (CMD); a value of 1 indicates balance (BLC), and the membership function is as follows. Figure 10c As shown.

[0117] Fuzzy rules:

[0118] Based on the average equivalent lateral swing angle The absolute value of the pilot's lateral control |δ lat,p |and its moving average Fuzzy rules are applied to infer the pilot's control intention. The fuzzy rules for the pilot's control intention are shown in Table 1 below. It should be noted that when the yaw angle is large, the goal is to stabilize the sling oscillation regardless of the amplitude of the pilot's input. When the pilot's instantaneous input is equal to the moving average input, the pilot expects the helicopter to exhibit balanced response characteristics (BLC).

[0119] Table 1

[0120]

[0121] Deblurring:

[0122] By applying fuzzy rules, variables can be determined. Given the current input values, the membership degree is obtained at this moment. As shown in the following formula.

[0123]

[0124] In the formula, N,S i ,v i These are the fuzzy level, membership degree, and fuzzy value, respectively.

[0125] Post-output processing:

[0126] To prevent To mitigate the sudden increase in adverse effects on the pilot, a gain-increasing damper, a first-order inertial element, was added after the fuzzy output to achieve... The slow increase. In contrast, if If the signal is reduced, it is provided directly to the controller without going through the damping circuit, so as to quickly achieve command tracking.

[0127] (4) Validation of the effectiveness of the improved sling swing angle feedback method based on fuzzy reasoning of pilot control intention

[0128] After improving the sling swing angle feedback, its effectiveness needs to be verified, and the parameters adjusted based on the response calculation results. For example, [the following method is used]. Figure 11The diagram shows the structure of the human-machine-sling coupled closed-loop system. The helicopter sling flight dynamics model and flight control system are as described above. The pilot model adopts Hess's structured pilot model (e.g., Figure 12 As shown in the figure, it includes a proprioceptive feedback loop, a vestibular feedback loop, and a visual feedback loop.

[0129] The control element is composed of Y C This indicates that, corresponding to the closed-loop control system composed of the helicopter sling-mounted flight dynamics model and the flight control system, Y NM Y represents the pilot's neuromuscular system. FS This refers to the force sensing system of a helicopter. Y PF Based on the pilot's proprioceptive characteristics, and the control stick deflection δ F Obtain the proprioceptive control quantity. Y VF The vestibular system characteristics of a pilot are used to sense acceleration or angular velocity and output a vestibular control quantity U. VF Y P This represents the pilot's outer-loop compensation, which determines the desired attitude C based on the position error, followed by inner-loop compensation Y. E After the central system time delay τ0, the pilot visual compensation control U based on attitude visual error is obtained. E Table 2 below shows the nominal values ​​of the above units, K. * This indicates that the gain of the unit can be adjusted according to different compensation tasks.

[0130] Table 2

[0131]

[0132] After establishing a human-machine-sling coupling closed-loop system, numerical simulation is performed using the mission element specifications in Helicopter Flight Quality Specification ADS-33E. The effectiveness of the human-machine cooperative fuzzy CAF is determined based on the simulation results, and the parameters are adjusted according to the response results. For example... Figure 13a , Figure 13b The paper presents numerical simulation results for lateral positioning based on a human-machine-suspended closed-loop system. The human-machine cooperative fuzzy CAF proposed in this embodiment is denoted as FUZZY-CAF. The simulation results are compared with those of two other CAFs (CT and SD). CT-CAF shows the best command tracking performance, while SD-CAF shows the best suspension swing damping. Figure 13b It can be determined that a small residual hanging oscillation remains under the CT-CAF configuration, while the residual oscillation disappears under the SD-CAF and FUZZY-CAF configurations. Figure 13aIt can be determined that the helicopter roll response under the SD-CAF configuration differs significantly from the expected response, while the responses under the CT-CAF and FUZZY-CAF configurations match the expected responses well. This indicates that FUZZY-CAF combines the advantages of CT-CAF's good command tracking performance and SD-CAF's fast sling swing convergence, reflecting the effectiveness of the real-time configuration of the sling swing angle feedback. If the FUZZY-CAF performance is found to be less than expected, it can be adjusted by adjusting the membership function of the sling lateral swing rate gain.

[0133] This invention provides a method for reducing sway in helicopter pilot control intention using fuzzy inference sling angle feedback. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment of the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A method for reducing the sway of a helicopter hoist cable with ambiguous pilot control intent feedback, comprising: The method comprises the following steps: Step 1, establishing a helicopter suspended load system (HSLS) model; Step 2, establishing a classical explicit model following (MFCS) control system integrated with a sling angle feedback control law (CAF) and a fuzzy module, coupling the helicopter suspended load system (HSLS) model to form a helicopter suspended load flight closed-loop system model; Step 3, detailed design of the fuzzy module, obtaining pilot manipulation intention by using fuzzy reasoning according to the suspended load flight state and pilot manipulation history; Step 4, adjusting the sling angle feedback control law configuration in real time according to the pilot manipulation intention to eliminate conflicts, and verifying the effectiveness by simulating the human-machine-suspended load coupled closed-loop system. In step 1, the inputs of the helicopter suspended load system (HSLS) model are rotor automatic inclinators deflection and tail rotor pitch, and the outputs are helicopter rigid body motion state and sling angle derived from the suspended load system.

2. The method of claim 1, wherein, In step 1, the flapping motion equation and the edgewise motion equation of the i-th blade are established by setting the rotor blade as a rigid body, accumulating the inertia load, aerodynamic load and damping moment of the blade to the balance at the hinge, as shown in the following formula: where ψ i is the azimuth angle of the rotor; β i , denote the flap angle, flap angular velocity and flap angular acceleration of the i-th blade piece, respectively; ζ i , denote the lag angle, lag angular velocity and lag angular acceleration of the i-th blade piece, respectively; e is the hinge offset; S b and I b denote the blade sectional moment of inertia and the blade sectional moment of inertia, respectively; Ω, denote the rotor rotational speed and rotational acceleration, respectively; is the three-axis acceleration of the hub; p s , q s , r s are the three-axis angular velocities of the hub, is the three-axis angular acceleration of the hub; and are the flap and lag aerodynamic moments, respectively; and are the lag damper-to-flap and lag damper-to-lag moments of action, respectively.

3. The method of claim 2, wherein, In step 1, the state space form of the helicopter suspended load system (HSLS) model is: where x and are the state vector of the HSLS model and its derivative, respectively; f is the state equation; δ is the control input vector, and t is time, as follows: delta = (delta col , delta lat , delta lon , delta tr ) T where v0, v 1C , 1S are rotor dynamic inflow state variables; v 0TR is the tail rotor uniform inflow state variable; u, v, w are the helicopter's three-axis velocity vector; p, q, r are the helicopter's three-axis angular velocity; φ, θ, ψ are the helicopter's three attitude angles; x, y, z are the helicopter's three-axis position in the ground axis system; u L , v L , w L are the sling's three-axis velocity vector; p L , q L , r L are the sling's three-axis angular velocity; φ L , θ L , ψ L are the sling's three attitude angles; x L , y L , z L are the sling's three-axis position in the ground axis system; δ col , δ lat , δ lon , δ tr represent the helicopter's collective, lateral cyclic, longitudinal cyclic and tail rotor collective, respectively; T represents the matrix transpose.

4. The method of claim 3, wherein, In step 1, the sling angle is determined by the relative motion between the helicopter and the suspended object, relative to the Earth's axis (O-XYZ). E A consistent hook coordinate system (O-XYZ) K The longitudinal swing angle α is defined in the middle. c and lateral swing angle β c , origin O K It is the average position of the suspension point on the helicopter, with the origin O. K On the body axis (O-XYZ) B The relative position in is defined as This indicates the position of the center of gravity of the suspended object in the hook coordinate system; to obtain the swing angle, Represented by the swing angle and HSLS state variables respectively, the following relationship is established: where A KB is the transformation matrix from the body axis to the hook coordinate system; l c is the distance from the average hook point on the helicopter to the center of gravity of the sling load.

5. The method of claim 4, wherein, In step 2, the transfer function G of the classical explicit model following (MFCS) control system is: where M represents the command model of the classical explicit model following MFCS control system, P -1 represents the inverse model of the classical explicit model following MFCS control system, H represents the feedback loop of the classical explicit model following MFCS control system; Pτ is the controlled object, where τ is the equivalent time delay of high order dynamics.

6. The method of claim 5, wherein, In step 2, under the requirement of ADS-33E-PRF, the ideal helicopter flight dynamics and flight qualities are obtained by designing the command model, the roll channel and the pitch channel adopt the attitude command and attitude holding response type, and the vertical channel and the yaw channel adopt the speed command and height or direction holding response type; The sling angle feedback loop L and the fuzzy module Z are added to the classical explicit model following (MFCS) control system; The helicopter suspended load system (HSLS) model, the classical explicit model following (MFCS) control system, the sling angle feedback loop and the fuzzy module are coupled to form a helicopter suspended load flight closed-loop system model.

7. The method of claim 6, wherein, In step 3, the structure of man-machine collaborative fuzzy sling angle feedback is established, the fuzzy module is designed in detail, and the longitudinal swing angle α c and angular velocity are fed back to the longitudinal channel, the lateral swing angle β c and angular velocity are fed back to the lateral channel; the sling swing angle needs to be processed by a high-pass filter to balance the steady-state swing angle, the fuzzy module receives the lateral swing angle β c , angular velocity and the pilot's lateral control δ lat,p , and obtains the time-varying lateral swing angle rate gain through fuzzy reasoning Specifically includes the following steps: Step 3-1, input pre-processing: an equivalent swing angle algorithm is proposed based on the undamped simple pendulum model, as shown below, β c and is processed as an equivalent lateral swing angle β c,e , that is, the maximum swing angle corresponding to the swing angle rate of zero: where m L is the mass of the suspended object, g is the acceleration due to gravity; Subsequently, β c,e The average equivalent lateral yaw angle over several seconds is obtained using a moving average algorithm, as shown in the following equation As the first input variable of the fuzzy inference system, the absolute value δ lat,p As the second input variable, the third input quantity is the moving average of the absolute value of the lateral control input where g represents gravity; T β and T lat represent the integration time of β c,e and the integration time of |δ lat,p | respectively; Step 3-2, Blurring: Blurring involves three blurring inputs: average equivalent lateral sway angle |δ lat,p | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average | and its moving average < Step 3-3, fuzzy rules: Based on the average equivalent lateral swing angle The absolute value of the lateral pilot control |δ lat,p | and its moving average The pilot control intention is inferred by applying fuzzy rules; Step 3-4, defuzzification: determining the variable by applying fuzzy rules the membership at the present input value, resulting in the where N, S i ,v i , respectively represent the fuzzy grade, membership degree and fuzzy value. Step 3-5, output post-processing: add a gain-up damping to the fuzzy output, which is a first order inertia link, to achieve slow increase, if decrease, then directly provide to the controller without damping link.

8. The method of claim 7, wherein, In step 4, the pilot, the control system and the helicopter suspended load system form a human-machine-suspended load coupled closed-loop system; The pilot model adopts the Hess's structured pilot model, including the proprioceptive feedback loop, the vestibular feedback loop and the visual feedback loop; The control element Y C represents the pilot's neuromuscular system, Y NM represents the pilot's neuromuscular system, Y FS represents the pilot's neuromuscular system, Y PF represents the pilot's neuromuscular system, Y F represents the pilot's neuromuscular system, Y VF represents the pilot's neuromuscular system, Y VF represents the pilot's neuromuscular system, Y P represents the pilot's neuromuscular system, Y E represents the pilot's neuromuscular system, Y E represents the pilot's neuromuscular system, Y After establishing the human-machine-suspended load coupled closed-loop system, the task subject primitives specified in the helicopter flight quality specification ADS-33E are used for numerical simulation, the effectiveness of the sling angle feedback real-time configuration is judged according to the simulation results, and the parameters are adjusted according to the response results.

9. A storage medium, characterized by The computer program or instructions stored therein, when executed, implement the method of any one of claims 1 to 8.

Images