Traction anti-slip control method for railway vehicle

Abstract

The invention provides an anti-slip control method for traction of a railway vehicle. The anti-slip control method comprises the following steps: estimating and updating uncertain parameters in a train traction process in real time by adopting a slope online estimation method of a least square method with a forgetting factor; identifying the current optimal slip rate according to the estimated parameters; and a slip rate closed-loop control method based on sliding-mode control is adopted, a current applicable anti-slip control strategy is determined according to the identified optimal slip rate, and therefore the stability and reliability of the train traction process are guaranteed. According to the method, sufficient adhesive force support is provided for traction of the train, when a wheel pair has a sliding dynamic state and the adhesive force fluctuates, sliding mode control can rapidly adjust traction output and restrain idling according to information such as the optimal slip rate, and therefore the stability and reliability of the train traction process are effectively guaranteed.

Description

Technical Field

[0001] This invention relates to the field of vehicle anti-slip control technology, specifically to a method for anti-slip control of rail vehicle traction. Background Technology

[0002] Wheel slippage primarily occurs when traction force exceeds adhesion force. Therefore, controlling the magnitude of traction force can prevent slippage. In other words, controlling traction acceleration based on the adhesion force can prevent slippage. Thus, slippage prevention mainly focuses on controlling traction force and improving and fully utilizing adhesion force. Currently, rail vehicle traction systems employ various criteria to determine slippage, primarily relying on speed difference, acceleration, slip ratio, and differential acceleration, with speed difference and acceleration being the most common. Slippage prevention control systems use changes in speed difference and acceleration to set thresholds and control traction force changes to control wheel slippage. Regardless of the criterion used, preventing wheel slippage and fully utilizing adhesion are the main objectives of slippage prevention control. However, sometimes, two slippage prevention control systems using the same criteria can have different effects, mainly due to differences in the selection of slippage prevention criterion parameters and the traction force control process.

[0003] The current anti-slip control parameters are mainly determined by logic threshold control, which is a relatively simple and effective control method. It only requires setting appropriate threshold values, typically fixed values. However, this fixed threshold method has its shortcomings. Taking speed difference as an example, if the speed difference standard is set too low, it will cause the anti-slip system to malfunction; but if the speed difference standard is set too high, it will lead to reduced sensitivity. If the speed difference standard is set according to the high-speed range, normal anti-slip function cannot be guaranteed at low speeds. Therefore, the speed difference standard cannot be a fixed value, but should be a function of speed. That is to say, the threshold based on speed difference should increase with the increase of train speed, and should be an uphill function that changes with the increase of train speed. During traction, as the speed increases, the adhesion coefficient decreases, and the acceleration of the vehicle during traction also decreases. The threshold for acceleration should also be a function that changes with the train speed, not a fixed value. The slip ratio is a function of speed difference and speed; therefore, when using slip ratio as a criterion, the selection of the threshold should also consider the influence of speed changes. Therefore, it is necessary to study the control parameters of commonly used criteria such as speed difference, acceleration, and slip ratio, and to find a new anti-skid control method and its corresponding criteria. Summary of the Invention

[0004] In view of the deficiencies in the prior art, the purpose of this invention is to provide a method for preventing slippage during traction of rail vehicles, which can effectively ensure the stability and reliability of the train traction process.

[0005] To solve the above problems, the technical solution of the present invention is as follows:

[0006] A method for preventing slippage during traction control of rail vehicles includes the following steps:

[0007] An online slope estimation method with forgetting factor and least squares is used to estimate and update uncertain parameters in the train traction process in real time.

[0008] Identify the current optimal slip ratio based on the estimated parameters;

[0009] A slip ratio closed-loop control method based on sliding mode control is adopted. Based on the identified optimal slip ratio, the applicable anti-slip control strategy is determined to ensure the stability and reliability of the train traction process.

[0010] Prior to this, the step of using the online slope estimation method with forgetting factor to estimate and update the uncertain parameters in the train traction process in real time specifically includes: firstly, performing online estimation of the slope of the stick-slip characteristic curve, obtaining real-time vehicle speed and wheel speed data during vehicle operation, obtaining the adhesion coefficient by combining the wheel-rail adhesion model, then obtaining the changing trend of the adhesion coefficient and slip ratio through data processing, dynamically estimating the slope of the stick-slip characteristic curve using the least squares method with forgetting factor, and using the slope of the stick-slip characteristic curve as a key indicator for identifying the optimal slip ratio.

[0011] Prior to this, the step of identifying the current optimal slip ratio based on the estimated parameters specifically includes: conducting optimal slip ratio identification, using the slope estimation result as the judgment criterion, initially setting a reasonable search step size and the assumed value of the optimal slip ratio; if the slope is positive, the slip ratio needs to be increased; if the slope is negative, the slip ratio needs to be decreased; at the same time, dynamically adjusting the search step size according to whether the slope sign changes abruptly, halving the step size when crossing the peak to avoid jitter during the search process, until the step size and slope fluctuations are both within a very small range, and determining the optimal slip ratio under the current operating conditions.

[0012] Prior to this, the step of employing a slip ratio closed-loop control method based on sliding mode control, which determines the applicable anti-wheel slip control strategy based on the identified optimal slip ratio to ensure the stability and reliability of the train traction process, specifically includes: using the slip ratio closed-loop control method based on sliding mode control, constructing a sliding mode control surface with the identified optimal slip ratio as the target value, designing an improved approaching law to reduce the common chattering problem in sliding mode control, deriving the traction force adjustment command in combination with the wheelset dynamic characteristics, and adjusting the traction force in real time by outputting the command through the traction control unit; when the actual slip ratio deviates from the target value, responding quickly and correcting to ensure that the slip ratio is stable in the optimal range and suppressing wheelset wheel slip.

[0013] Compared with existing technologies, this invention employs an online slope estimation method with a forgetting factor and least squares to estimate and update uncertain parameters in the train traction process in real time. The estimated parameters are used to identify the current optimal slip ratio. Then, a slip ratio closed-loop control method based on sliding mode control is used to recalculate the train's optimal adhesion. This invention's wheel-rail optimal adhesion online estimation algorithm can serve as a precise input for train traction adhesion enhancement control, enabling better implementation of anti-slip control, fully exploring the adhesion mechanism, and formulating corresponding adhesion control strategies. This provides sufficient adhesion support for train traction. When wheelsets exhibit slippage or adhesion fluctuates, sliding mode control can quickly adjust the traction output based on information such as the optimal slip ratio to suppress slippage, thereby effectively ensuring the stability and reliability of the train traction process. Attached Figure Description

[0014] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0015] Figure 1 This is a flowchart of the anti-slip control method for rail vehicle traction according to the present invention;

[0016] Figure 2 This is a force analysis diagram of the wheelset under traction conditions;

[0017] Figure 3 This is a schematic diagram illustrating the adhesion utilization of low-adhesion rail surfaces.

[0018] Figure 4 A schematic diagram illustrating the ultra-adhesion application of ultra-low adhesion rail surfaces;

[0019] Figure 5 This is a schematic diagram illustrating the adhesion of a variable adhesive track surface. Detailed Implementation

[0020] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0021] This invention addresses traction anti-slip control for rail transit vehicles, focusing on "real-time identification of the optimal slip ratio and precise control of the slip ratio." It achieves this through a three-step method: "online estimation of the stick-slip characteristic slope—optimal slip ratio identification—closed-loop adjustment of sliding mode control." Specifically, this invention provides a traction anti-slip control method for rail transit vehicles, such as... Figure 1 As shown, the method includes the following steps:

[0022] S1: The slope online estimation method with forgetting factor least squares is used to estimate and update the uncertain parameters in the train traction process in real time;

[0023] Specifically, the slope of the stick-slip characteristic curve is first estimated online. Real-time vehicle speed and wheel speed data during vehicle operation are obtained using dynamic software. The adhesion coefficient is obtained by combining the wheel-rail adhesion model (Polach model). Then, the changing trends of the adhesion coefficient and slip ratio are obtained through data processing. The slope of the stick-slip characteristic curve is dynamically estimated using the least squares method with a forgetting factor. This method reduces the influence of old data on the current estimation results by using the forgetting factor, which is more in line with the time-varying characteristics of the adhesion state during actual traction and provides an accurate basis for subsequent identification of the optimal slip ratio.

[0024] Identifying the slope of the stick-slip characteristic curve requires the current slip ratio of the wheelset and the adhesion as input. Currently, real vehicles can collect axle speed data in real time to calculate the slip ratio. In the stick-slip characteristic curve, the adhesion coefficient μ is the ordinate, and the slip ratio s is the abscissa (w is the wheel angular velocity). The curve slope k can be expressed as:

[0025]

[0026] In equation (4.1), when the slope is positive, the wheel-rail system is in a stable state, and increasing the slip ratio can still increase it using the viscosity coefficient. When the slope is negative, the wheel-rail system is in an unstable state; maintaining the current traction force will increase the wheelset slip ratio, decrease the viscosity coefficient, and pose a risk of wheelset spin. When the slope is zero, the viscosity coefficient reaches its maximum value, and the slip ratio at this point is the optimal slip ratio. Therefore, the slope of the viscosity-slip characteristic curve can be used as a key indicator for identifying the optimal slip ratio.

[0027] The process of estimating the slope using the least squares method is as follows:

[0028] According to equation (4.1), it can be known that make but Define the slope estimate as The estimated output for step n is:

[0029]

[0030] The deviation between the actual output and the estimated output is Then, in the estimation process from step 1 to step N, the performance index of the cumulative error can be defined in the following form:

[0031]

[0032] In the formula, Y=[y(1),y(2),…y(N)] T ,

[0033] The slope estimate that makes equation (4.3) reach its minimum value This is the desired least squares estimate. Regarding equation (4.3) with respect to... Taking the partial derivative, we can obtain the following expression:

[0034]

[0035] Obviously, when the slip ratio is not constant It is not always zero. Setting equation (4.4) to zero yields the estimated slope value:

[0036]

[0037] because Second-order partial derivative like If equation (4.5) can be satisfied, then J can achieve a minimum value.

[0038] Since Y = [y(1), y(2), ... y(N)] T , Equation (4.5) is used to calculate the slope estimate. When relying on data from the first to the nth step, the computational complexity and storage requirements are high. To ensure real-time computational efficiency, it is usually rewritten in a recursive least squares form.

[0039] Define the estimated value at step n as Represented in recursive form as Where q n This is the correction term. From equation (4.5), the slope estimate for the nth step can be obtained. for:

[0040]

[0041] In the formula P can be expressed as the sum of the data from the first n-1 steps and the data from the nth step. n Then there is

[0042]

[0043] After sorting, we can obtain

[0044]

[0045] From equation (4.6), we can obtain the estimated value for the (n-1)th step. for

[0046]

[0047] Substituting equation (4.8) into equation (4.9) yields

[0048]

[0049] From equations (4.6) and (4.10), the estimated value for the nth step is...

[0050]

[0051] Equations (4.11) and (4.7) can form the recursive formula for the least squares method.

[0052]

[0053] Since the slope of the stick-slip characteristic curve is usually time-varying during the anti-slip process, a forgetting factor ρ (usually taken as 0 < ρ ≤ 1) is introduced to enable the recursive least squares algorithm to effectively track the time-varying parameters. Combining with equation (4.3), the performance index J is redefined as:

[0054]

[0055] The forgetting factor ρ is used to reduce the influence of old data on the current estimate, making the algorithm focus more on recent data, thereby improving its ability to track time-varying parameters. Based on this performance metric, a recursive least squares algorithm with a forgetting factor can be derived similarly to equation (4.12) as follows:

[0056]

[0057] S2: Identify the current optimal slip ratio based on the estimated parameters;

[0058] Specifically, the optimal slip ratio is identified using the slope estimation result as the criterion. An initial reasonable search step size and optimal slip ratio assumption are set. If the slope is positive, it indicates that the current slip ratio is not optimal and needs to be increased; if the slope is negative, the slip ratio needs to be decreased. Simultaneously, the search step size is dynamically adjusted based on whether the slope sign changes abruptly (i.e., whether it crosses the adhesion peak). The step size is halved when crossing the peak to avoid jitter during the search process, until both the step size and slope fluctuations are within a minimal range, thus determining the optimal slip ratio under the current operating conditions.

[0059] Based on the slope identification method proposed above, the slope of the curve showing the change in adhesion coefficient and slip ratio can be obtained online. Then, an iterative search method can be used to identify the optimal slip ratio. That is, using the optimal slip ratio of the current step as a benchmark, if the slope value is positive, the optimal slip ratio of the next step will be larger; if the slope value is negative, the optimal slip ratio of the next step will be smaller. The recursive formula is as follows:

[0060] s op (i+1)=s op(i)+q(i)sgn(k(i)) (4.15)

[0061] In the formula s op q represents the optimal slip ratio, i is the number of iterations, k is the slope of the stick-slip characteristic curve, and q is the search step size.

[0062] To reduce oscillations and improve convergence efficiency during the iterative search process, the search step size q should be dynamically updated at each step. Specifically, the step size q is halved whenever the search process crosses a stickiness peak. This dynamic step size adjustment method reduces oscillations as the search approaches the optimal slip ratio while simultaneously accelerating the search. The update rules are as follows:

[0063]

[0064] S3: A slip ratio closed-loop control method based on sliding mode control is adopted. Based on the identified optimal slip ratio, the applicable anti-free-run control strategy is determined to ensure the stability and reliability of the train traction process.

[0065] Specifically, a slip ratio closed-loop control method based on sliding mode control is used to construct a sliding mode control surface with the identified optimal slip ratio as the target value. An improved approaching law is designed to reduce the chattering problem common in sliding mode control. The traction force adjustment command is derived by combining the wheelset dynamic characteristics. The traction control unit outputs the command to adjust the traction force in real time. When the actual slip ratio deviates from the target value, it responds quickly and corrects the error to ensure that the slip ratio is stable in the optimal range and to suppress wheelset idling.

[0066] Figure 2 This is a simplified diagram of the force analysis of a single wheelset during traction. During train traction, the traction torque (T) applied to the wheelset axle by the traction device generates a traction force F. t This traction force causes creep between the wheel and rail, generating a forward tangential force on the wheel-rail contact surface, which is the adhesive force F that causes the train to accelerate. μ Traction force F t It can be calculated using equation (4.17):

[0067]

[0068] Depend on Figure 2 The equation of motion for the wheelset rotation can be obtained as follows:

[0069]

[0070] In the formula: I w ω is the wheelset's moment of inertia; r is the wheelset's angular velocity during braking; ω is the wheelset's rotational inertia. w F is the wheelset radius; μ For wheel-rail adhesion; F t The braking force generated by the braking device.

[0071] From the wheel-rail dynamics analysis, the derivative of the traction wheel-rail slip ratio has the following form:

[0072]

[0073] Substituting the wheelset coupling dynamics equation (4.19) into (4.20), we get...

[0074]

[0075] With s tar As the target slip ratio, in order to make the slip ratio s converge to the target value s tar The following control strategy can be designed: when s > s tar When, let the derivative of the slip ratio To reduce the slip ratio toward the target value; when s tar season Increase the slip ratio toward the target value; when s = s tar season Maintain a stable slip ratio. Furthermore, when the actual slip ratio s deviates from the target value s... tar When the difference is large, the rate of change of the slip ratio should accelerate, that is... The absolute value is relatively large; while when s approaches s tar At that time, the rate of change should slow down, that is The absolute value is relatively small. Based on this, the control law can be designed as follows:

[0076]

[0077] From equations (4.16) and (4.17), we can obtain

[0078]

[0079] The above control strategy utilizes the concept of sliding mode control. To implement sliding mode control, firstly, the sliding surface needs to be determined, which defines the boundary where control signals switch during system control. Since the control objective is the wheelset slip ratio, the state variables x1 = v and x2 = ωr can be selected. w ,according to The state equations for the rotational motion of the wheelset are obtained.

[0080]

[0081] At this point, we have:

[0082]

[0083] Based on equation (4.15), the selectable sliding surface is: ​

[0084] S=ωr w -(1+s tar )v (4.26)

[0085] In the formula, s tar The target slip ratio.

[0086] Differentiating both sides of equation (4.16) simultaneously, we get:

[0087]

[0088] make Substituting equation (4.24) into equation (4.27), we get:

[0089]

[0090] The following equivalent control expression is obtained:

[0091]

[0092] Sliding mode control typically employs a constant velocity reaching law to enable the system to reach the sliding surface. The constant velocity reaching law is as follows:

[0093]

[0094] In the formula, γ is the slip gain and is greater than 0, and sgn represents the sign function.

[0095]

[0096] The presence of a sign function can easily induce chattering near the sliding surface, mainly due to the discontinuous nature of the sign function. Therefore, the sign function can be replaced with a continuous function to reduce chattering. The continuous sigmoid function has the following form:

[0097]

[0098] In the formula, λ>0.

[0099] To increase the speed of approaching the sliding surface, the constant velocity approach law can be replaced with an exponential approach law, and the sliding control gain can be increased. However, increasing the gain may lead to a large overshoot when reaching the sliding surface. Therefore, a differential term is introduced into the exponential approach law for optimization, resulting in an improved exponential approach law.

[0100]

[0101] In the formula, γ1, γ2, and γ3 are sliding mode gains, and all of them are greater than 0.

[0102] F in equation (4.19) μThis is an unknown quantity. Because sliding mode control is robust, even if F... μ While there is some deviation, the slip ratio can still be controlled near the target value. However, accurately obtaining this value would help improve control performance. The estimated adhesion force is defined as... F can be μ There exists an estimated value. but

[0103]

[0104] To ensure the effectiveness of the sliding mode control method, the accessibility of the sliding surfaces needs to be proven. To guarantee the accessibility of the switching surfaces, it is necessary to prove... According to equations (4.28) and (4.34), we can obtain

[0105]

[0106] Since all parameters are within a reasonable range, Therefore, the derivative term of the target slip ratio can be ignored, thus simplifying equation (4.35) to:

[0107]

[0108] Organized

[0109]

[0110] When s≠s tar When S≠0, then we have Substituting it into equation (4.37) yields

[0111]

[0112] Therefore, the reachability proof can be transformed into a proof that equation (4.38) holds:

[0113]

[0114] From equation (4.39), it can be seen that it is only necessary to satisfy... This guarantees that the expression holds true. That is, the sliding surface can be reached in a finite amount of time.

[0115] When s = s tar When S = 0, Therefore, the reachability of the switching surface can be satisfied.

[0116] Effect verification:

[0117] Figure 3 This is a schematic diagram illustrating the adhesion utilization of low-adhesion rail surfaces, such as... Figure 3As shown, the actual adhesion utilization value (red curve) fluctuates very little on the adhesion characteristic surface, almost perfectly matching the distribution of the peak trajectory (blue curve) of the surface. This phenomenon indicates that the control strategy proposed in this invention can achieve efficient capture and stable utilization of the maximum adhesion resources under low adhesion conditions, and there is no significant shift or oscillation of the adhesion coefficient on both sides of the characteristic surface, which fully verifies its technical advantages in adhesion utilization accuracy and control stability.

[0118] Figure 4 This is a schematic diagram illustrating the ultra-adhesion application of ultra-low adhesion rail surfaces, such as... Figure 4 As shown in the adhesion characteristic surface, the degree of adhesion utilization is very high at this time, which verifies and illustrates that even under ultra-low adhesion conditions, the control strategy proposed in this invention still has a good control effect.

[0119] Figure 5 This is a schematic diagram illustrating the adhesion application of a variable adhesive surface, as shown below. Figure 5 As shown, during actual vehicle operation, low adhesion can often be sudden due to oil stains or leaves adhering to a certain section of the track, thus the adhesion state may change abruptly. The train travels on a dry track at speeds below 100 km / h, then on a low-adhesion track at speeds of 100–200 km / h, and then abruptly transitions to ultra-low adhesion at speeds of 200–300 km / h. It can be seen that the online adhesion estimation can be rapidly adjusted according to the changes in track surface condition, achieving dynamic online adhesion estimation. On a dry track, the adhesion utilization is a straight line on the left side of the curved surface. At this point, the online adhesion estimation only yields the utilization of adhesion corresponding to the current train acceleration, not the maximum track surface adhesion. However, once the wheels spin (track surface abrupt change), the adhesion utilization in the spinning state only fluctuates at the instant of the adhesion change and completes adaptive adjustment of the track surface within approximately 2 seconds. The degree of adhesion utilization is very high, almost reaching the maximum adhesion state of the track surface. Therefore, the control strategy proposed in this invention can control the track surface adhesion utilization at the optimal adhesion state.

[0120] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for preventing slippage during traction control of rail vehicles, characterized in that, The method includes the following steps: An online slope estimation method with forgetting factor and least squares is used to estimate and update uncertain parameters in the train traction process in real time. Identify the current optimal slip ratio based on the estimated parameters; A slip ratio closed-loop control method based on sliding mode control is adopted. Based on the identified optimal slip ratio, the applicable anti-slip control strategy is determined to ensure the stability and reliability of the train traction process.

2. The method for preventing slippage of rail vehicle traction according to claim 1, characterized in that, The steps of using the online slope estimation method with forgetting factor to estimate and update uncertain parameters in the train traction process in real time include: firstly, online estimation of the slope of the stick-slip characteristic curve is performed, real-time vehicle speed and wheel speed data during vehicle operation are obtained, the adhesion coefficient is obtained by combining the wheel-rail adhesion model, and then the changing trend of the adhesion coefficient and slip ratio is obtained through data processing. The slope of the stick-slip characteristic curve is dynamically estimated using the least squares method with forgetting factor, and the slope of the stick-slip characteristic curve is used as a key indicator for identifying the optimal slip ratio.

3. The method for preventing slippage during traction control of rail vehicles according to claim 1, characterized in that, The step of identifying the current optimal slip ratio based on the estimated parameters specifically includes: conducting optimal slip ratio identification, using the slope estimation result as the judgment standard, initially setting a reasonable search step size and the assumed value of the optimal slip ratio; if the slope is positive, the slip ratio needs to be increased; if the slope is negative, the slip ratio needs to be decreased; at the same time, dynamically adjusting the search step size according to whether the slope sign changes abruptly, halving the step size when crossing the peak to avoid jitter during the search process, until the step size and slope fluctuations are both within a very small range, and determining the optimal slip ratio under the current operating conditions.

4. The method for preventing slippage during traction control of rail vehicles according to claim 1, characterized in that, The steps of employing a slip ratio closed-loop control method based on sliding mode control, which determines the applicable anti-wheel slip control strategy based on the identified optimal slip ratio to ensure the stability and reliability of the train traction process, specifically include: using the slip ratio closed-loop control method based on sliding mode control, constructing a sliding mode control surface with the identified optimal slip ratio as the target value, designing an improved approaching law to reduce the chattering problem common in sliding mode control, deriving traction force adjustment commands based on wheelset dynamic characteristics, and adjusting the traction force in real time by outputting commands through the traction control unit; when the actual slip ratio deviates from the target value, responding quickly and correcting to ensure that the slip ratio is stable within the optimal range and suppressing wheelset wheel slip.

Images