CN-122008889-A - Designated performance control method for preventing adsorption deadlock during climbing of high-speed maglev train

Abstract

The invention discloses a designated performance control method for preventing adsorption deadlock during climbing of a high-speed maglev train, and belongs to the field of maglev rail transit. The method adopts a dynamic expansion asymmetric preset performance constraint and generalized model predictive control technology, takes RBFNN as a model mismatch controller, performs safe small air gap suspension control for preventing adsorption deadlock on a maglev train system, establishes a high-speed maglev train, designs the dynamic expansion asymmetric preset performance constraint by the stagger track model, establishes a high-speed maglev train suspension air gap mathematical model under the preset performance constraint, designs a generalized model predictive controller under the dynamic expansion asymmetric preset performance constraint for preventing the problems of excessive energy loss input by the system deadlock and the controller during climbing of the train, adopts a model mismatch control approximate high-speed maglev train suspension system unknown item based on a radial basis neural network, and ensures that the system can still safely and stably run when facing the problems of external interference, stagger tracks and the like while improving the low-power consumption small air gap suspension of the maglev train.

Inventors

• WANG WENJUN
• CHU XIAOGUANG
• KONG YING

Assignees

• 曲阜师范大学

Dates

Publication Date

20260512

Application Date

20260401

Claims (5)

1. 1. The method for controlling the specified performance of the high-speed maglev train in climbing and preventing adsorption deadlock is characterized by comprising the following steps of: s1, establishing a staggered track model of a high-speed maglev train, wherein the staggered track model comprises the following steps: (1) Wherein, the Is the speed of the high-speed magnetic levitation train, Are all different track staggering heights, Is that Is provided with a track span of (a), Is that Is a track span of (2); Based on the availability of (1), the air gap reference tracking of the high-speed maglev train is as follows: (2) Wherein, the The actual expected suspension air gap for the high-speed magnetic levitation train; S2, designing anti-adsorption deadlock dynamic expansion asymmetric preset performance constraint containing staggered track information; S3, constructing a high-speed magnetic levitation train suspension air gap mathematical model under the constraint of preset performance; s4, designing a generalized model prediction controller under dynamic expansion asymmetric preset performance constraint; S5, designing a model mismatch controller based on the radial basis function neural network.
2. 2. The method for controlling the specified performance of the anti-adsorption deadlock for the climbing of the high-speed maglev train according to claim 1, wherein the specific steps of designing the anti-adsorption deadlock dynamic expansion asymmetric preset performance constraint containing staggered track information in the step S2 are as follows: s21, designing fixed upper and lower boundaries containing staggered platform track information , The method comprises the following steps: (3) Wherein, the To fix the initial values of the upper and lower boundaries, For a steady state value that fixes the upper and lower boundaries, l 1 ,l 2 is an adjustable constant; s22, designing preset upper and lower boundaries of preset performance constraint , The method comprises the following steps: (4) Wherein, the H max is the maximum gap between the wheel tracks of the maglev train; S23, designing an expansion mechanism for presetting the upper and lower boundaries of performance constraint , The method comprises the following steps: (5) Wherein the error is H is the actual expected suspension air gap of the high-speed magnetic levitation train; S24, designing an effective expansion mechanism of the upper boundary and the lower boundary of a preset performance constraint expansion function , The method comprises the following steps: (6) Wherein, the Is an adjustable constant; S25, designing an expansion function of the upper boundary and the lower boundary of the preset performance constraint according to formulas (4), (5) and (6) , The method comprises the following steps: (7) Wherein, l U ,l L is an adjustable constant; s26, in conclusion, designing a preset performance constraint function for dynamically expanding asymmetric upper and lower boundaries , The method comprises the following steps: (8)。
3. 3. the method for controlling the specified performance of the high-speed maglev train climbing anti-adsorption deadlock according to claim 2, wherein the specific steps of constructing the high-speed maglev train levitation air gap mathematical model under the preset performance constraint in the step S3 are as follows: s31, constructing a high-speed maglev train suspension air gap mathematical model: (9) Wherein m is the mass of the bogie, g is the gravitational acceleration, h is the upward floating height of the suspension train, For the upward floating speed of the train, F is a levitation magnetic force, F 0 is a disturbance between representing multiple subsystems, usually zero, F d is an instantaneous external disturbance such as aerodynamic lift and levitation quality change unless mechanical decoupling fails, i is levitation current (replaced by control input U below), mu 0 is vacuum permeability, N is coil turns, A is magnetic pole area, H 0 is the maximum levitation air gap of the magnetic levitation train, U dc is the power supply voltage of an H bridge driving circuit, L is the equivalent inductance of a levitation coil, R is the equivalent resistance of the levitation coil, ; S32, obtaining the linear state space equation of the high-speed maglev train suspension system by the formula (9) (10) Where u r is the reference control input, Including linearization errors and system uncertainty terms, , ; S33, converting error under preset performance constraint The state equation for a state vector is: (11) Wherein, the , , , , , , For the actual expected levitation acceleration of the high-speed maglev train, Is the first-order derivative of the error, , The method is used for dynamically expanding preset performance constraint functions of asymmetric upper and lower boundaries.
4. 4. The method for controlling the specified performance of the high-speed maglev train climbing anti-adsorption deadlock according to claim 3, wherein the generalized model predictive controller under the constraint of dynamically expanding asymmetric preset performance is designed in the step S4, and the specific steps are as follows: s41 design based on reconstruction error The prediction model of (2) is: (12) Wherein, the , In order to predict the time domain of the signal, Is comprised of Higher-order terms of (2); s42, designing a cost function by dynamically expanding an asymmetric preset performance function, wherein the cost function is as follows: (13) S43, dynamically expanding the generalized model predictive controller under the asymmetric preset performance constraint to solve the optimal solution The method comprises the following steps: (14)。
5. 5. The method for controlling the specified performance of the high-speed maglev train climbing anti-adsorption deadlock according to claim 4, wherein the designing of the model mismatch controller based on the radial basis function neural network in S5 comprises the following specific steps: s51, optimal control input including RBFNN approximation: (15) Wherein, the Is that Is a function of the estimated value of (2); s52, substituting (15) into the state space equation (11) can obtain: (16) Wherein, the ; S53, approximating the unknown item with RBFNN : (17) Wherein, the Is that Estimating matrix, input vector Z, weight vector Z The basis function vector is , As weights, r m is a basis function, where ; S54, designing a Gaussian basis function as follows: (18) Wherein c ij and b i respectively represent the ith error vector [ ] ) The center vector and width of the j-th neuron in (a), the subscript j denotes the location of the neuron in the network, in particular ; S55 substituting (17) into (15), the following can be obtained: (19) s56, designing a self-adaptive updating rule of RBFNN pair weights in a high-speed maglev train suspension system Can be expressed as: (20) Wherein, the Is a combination of two positive coefficients, the two positive coefficients, Is a positive definite matrix; S57: With optimal weight The method comprises the following steps: (21) wherein the model approximation error can be expressed as , As an approximation error of the network, Is ideal for 。

Description

Designated performance control method for preventing adsorption deadlock during climbing of high-speed maglev train Technical Field The invention relates to the field of magnetic levitation trains, in particular to a specified performance control method for preventing adsorption deadlock during climbing of a high-speed magnetic levitation train. Background The high-speed magnetic levitation train is gradually establishing the strategic position in the next generation track traffic system by virtue of the high running speed, the excellent riding comfort and the low-carbon environment-friendly characteristics of the high-speed magnetic levitation train. However, under complex line conditions, particularly in the operating conditions of a variable-gradient road section, how to ensure the dynamic stability and the operation safety of the train in the climbing process has become one of the key technical bottlenecks restricting the engineering application of the train. When a train drives into a ramp section, the train body generates a remarkable pitching motion mode due to geometric linear change of the track, so that an air gap between the suspension electromagnet and the track is subjected to severe transient change at the front end or the rear end. If the suspension control strategy can not realize quick and robust response to the dynamic disturbance, the phenomenon of vehicle body adsorption deadlock is extremely easy to be induced, namely the front suspension air gap is increased sharply, so that the electromagnetic suction force is reduced below a critical value, and further suspension failure and even mechanical contact risks are caused. This phenomenon not only severely degrades ride comfort indicators, but also is more likely to trigger a system protective shutdown mechanism, resulting in operational disruption. Notably, the challenges presented by trains traveling on ramps such as periodic rough tracks, off-track tracks, and the like are far more than these. When the train passes through the slope change point, the abrupt change of the track curvature can cause nonlinear disturbance of the suspension system, and the pitching oscillation of the train body is aggravated. Meanwhile, the vehicle needs larger traction force in the climbing process, and the electromagnetic coupling effect between the traction system and the suspension system is further enhanced, so that the dynamic environment of suspension control is more complex. In addition, factors such as track irregularity, uneven load distribution, wind resistance interference and the like can be amplified on a ramp section, and higher requirements are put on the disturbance rejection capability of a suspension system. Due to the inherent open loop instability of a suspension system of the maglev train and the strong nonlinearity and time-varying characteristics of a system dynamics model under a ramp section, accurate constraint on the posture (particularly the pitching angle) of a train body in the climbing process is realized, and the head-up trend is restrained, so that the maglev train has become a very challenging technical problem. Therefore, the invention provides a high-speed maglev train climbing anti-adsorption deadlock control method based on preset performance constraint, which aims to ensure that a train can still maintain a stable suspension posture on a high-gradient road section through active constraint and dynamic compensation of the posture of a train body, so that the full-working-condition safe and stable operation under the condition of a complex line is ensured. Disclosure of Invention Aiming at the defects and the blank of the prior art, the invention provides the designated performance control method for preventing the adsorption deadlock of the climbing of the high-speed maglev train, which improves the sensitivity of the high-speed maglev train to the position error by dynamically and rapidly expanding the control method for the small air gap levitation of the high-speed maglev train under the constraint of preset performance and limits the high-speed maglev train to be in the set constraint boundary, so that the high-speed maglev train is in the mechanical boundary, the deadlock breakdown of the system is avoided, and the safe operation of the high-speed maglev train under the small air gap is ensured. In order to achieve the purpose, the invention provides a specified performance control method for preventing adsorption deadlock during climbing of a high-speed maglev train, which comprises the following steps: Step 1, establishing a staggered track model of a high-speed magnetic levitation train; step 2, designing anti-adsorption deadlock dynamic expansion asymmetric preset performance constraint containing staggered track information; Step 3, constructing a high-speed magnetic levitation train suspension air gap mathematical model under the constraint of preset performance; step 4, designing a generalized model predictio