Vehicle power adaptive allocation and cooperative control method based on time sensitive network

Abstract

The application provides a vehicle power adaptive distribution and cooperative control method based on a time-sensitive network, designs a vehicle control architecture based on the time-sensitive network, the central decision unit collects the load, temperature and health state of each local execution unit through TSN synchronization, introduces dynamic constraints such as wheel-rail adhesion to perform multi-constraint adaptive optimization solving, realizes global optimal distribution of the traction force and the electric braking force of each unit, adopts a prediction-dynamic compensation mechanism in the braking working condition, perceives the electric braking capacity attenuation in advance and smoothly introduces air brake compensation, uses the TSN deterministic time window to realize impact-free synchronous switching and closed-loop fine adjustment of the double-braking instruction, breaks the information island of the traditional network and eliminates the communication delay uncertainty, realizes global optimal distribution and accurate cooperative control of the whole vehicle power resource, effectively suppresses the braking impact and improves the ride comfort.

Description

Technical Field

[0001] This invention relates to the field of rail transit technology, and in particular to a method for adaptive allocation and cooperative control of vehicle power based on time-sensitive networks. Background Technology

[0002] Currently, subway vehicles generally adopt distributed control systems based on traditional networks (such as MVB), which suffer from the following long-standing technical bottlenecks that have not been properly resolved: The "information silos" lead to suboptimal control issues: Traction, electric braking, and air braking systems typically have independent control units, resulting in high information exchange delays and data asynchrony between them. This prevents the global optimal allocation of power resources at the vehicle level. For example, during braking, the electric braking system cannot perceive the adhesion status and load of each drive axle in real time, potentially leading to uneven distribution of electric braking force, affecting adhesion utilization and energy efficiency. Furthermore, the switching between electric and air braking relies on simple threshold judgments, lacking a smooth transition mechanism, causing braking shocks and impacting comfort.

[0003] Communication uncertainty leads to control performance bottlenecks: Traditional network communication delays exhibit random jitter, preventing the power units in each carriage from executing commands at precisely the same moment. For cooperative control requiring high synchronization (such as electro-pneumatic combined braking), this slight asynchrony directly translates into longitudinal shock, reducing ride quality. Existing technologies mitigate this problem by adding buffers at the mechanical or hydraulic levels, but these methods are costly and have limited effectiveness.

[0004] The problems of system rigidity and high life cycle costs: The existing vehicle network and control logic are tightly coupled. When the train formation changes or new intelligent components are added, a large amount of hardware wiring, network configuration and software debugging need to be redone. The system has poor scalability and flexibility, and the maintenance and upgrade costs are high.

[0005] Although TSN technology is considered the future of industrial communication due to its deterministic latency, ultra-high reliability (frame redundancy), and time synchronization (IEEE 802.1AS), its current applications in rail transit are mostly limited to replacing the communication network itself and designing hardware boards, or only used for non-safety-related data transmission (such as video surveillance). There is currently no publicly available technical solution that deeply integrates TSN's communication performance into the core power control loop of a vehicle, and forms a complete method to systematically solve the aforementioned problems. Summary of the Invention

[0006] In view of the above problems, the present invention is proposed to provide a time-sensitive network-based vehicle dynamics adaptive allocation and cooperative control method that overcomes or at least partially solves the above problems.

[0007] This invention discloses a vehicle power adaptive allocation and cooperative control method based on time-sensitive networks, the method comprising: The central decision-making unit is connected to multiple execution units through a communication backbone based on Time Sensitive Network (TSN), and the clocks between the central decision-making unit and each execution unit are synchronized using the TSN time synchronization protocol. The central decision-making unit periodically obtains the total power demand command of the vehicle and the real-time status information fed back by each execution unit. The real-time status information includes at least the load information of the compartment corresponding to each execution unit, the temperature information of each execution unit, and the health status information. Based on the received real-time status information, the central decision-making unit determines the dynamic capability constraints of each execution unit under the current operating conditions. Under the premise of satisfying the total power demand constraints of the vehicle, the dynamic capability constraints of each execution unit, and the load constraints based on load information, the central decision-making unit calculates the power distribution command of each execution unit through a preset optimization allocation algorithm. In the braking condition, it also includes a cooperative constraint so that the required air braking force is determined based on the comparison between the total braking demand and the total available electric braking capacity of each electric braking unit. The central decision-making unit adopts the time-aware shaping mechanism of the TSN network to reserve a transmission time window with deterministic delay for power allocation commands, and synchronously sends each power allocation command to the corresponding execution unit within the same control cycle; Each execution unit performs traction or braking actions according to the received power distribution command and feeds back the execution status, including the actual output value, to the central decision-making unit.

[0008] Optionally, the TSN time synchronization protocol is used to synchronize the clocks between the central decision-making unit and each execution unit, including: The IEEE 802.1AS protocol is used to synchronize the clocks between the central decision-making unit and each execution unit.

[0009] Optionally, load information is obtained by real-time calculation of the pressure signals of the air springs in each carriage, temperature information is obtained by real-time acquisition of the temperature sensors of the motors or inverters in each execution unit, and health status information is determined comprehensively based on the fault diagnosis information fed back by each execution unit.

[0010] Optionally, the dynamic capability constraints of each execution unit under the current operating conditions include: The maximum traction or braking force of each actuator under the current operating conditions is limited by the temperature and health status information of the actuator, and also by the wheel-rail adhesion conditions determined based on the real-time vehicle speed and wheel-rail condition. The minimum traction force or minimum braking force of each actuator is limited by the reverse allowable ratio of the maximum traction force or maximum braking force of each actuator under the current operating conditions.

[0011] Optionally, the preset optimized allocation algorithm: Under traction conditions, the traction force of each actuator is distributed with the goal of minimizing the total energy consumption of the vehicle. Under braking conditions, the distribution of electric braking force and air braking force among the actuators is prioritized with maximizing the energy recovery of electric braking while also taking into account braking comfort.

[0012] Optionally, The preset optimization allocation algorithm, under braking conditions, takes maximizing the sum of available electric braking capabilities provided by each electric braking control unit as the optimization objective, and introduces a comfort weight coefficient calibrated through a real vehicle to constrain the deceleration rate of change, so as to adjust the allocation relationship between electric braking and air braking.

[0013] Optionally, the method further includes: When the central decision-making unit has limited computing resources, the optimized allocation algorithm is replaced with a fast allocation method based on fuzzy rules or a preset MAP. The fuzzy rules take the load deviation, temperature deviation and health status of each execution unit as input variables and output the correction amount of the allocation coefficient. The preset MAP stores the allocation coefficient with load and temperature as dimensions and calculates the real-time allocation coefficient through an interpolation algorithm.

[0014] Optionally, the method further includes: During braking, the central decision-making unit continuously monitors the real-time status of each electric braking control unit. When it is predicted that the electric braking capacity of any electric braking control unit will decrease due to temperature rise, a transition command including gradual reduction of electric braking force and gradual increase of air braking force is generated in advance to control the smooth switching of the electric braking system and the air braking system. The electric braking force command and the air braking force command in the transition command are generated using an S-shaped transition curve with continuous derivatives to ensure that the total braking force remains continuous throughout the transition period, while the deceleration rate of change is maintained within the preset comfort range. After a smooth transition, the central decision-making unit compares the deviation between the actual output force fed back by each electric braking control unit and the corresponding command value with the actual output force fed back by each air braking control unit. In the next control cycle, it dynamically compensates and adjusts the power distribution command of each execution unit according to the deviation to ensure the stability of the total braking force of the vehicle.

[0015] Optionally, the method further includes: An extended Kalman filter algorithm is used to predict the temperature change trend after a preset time by using the temperature and health status of the electric braking control unit as state variables, and to calculate the corresponding predicted value of the available electric braking capacity based on the attenuation law calibrated by the component characteristics.

[0016] Optionally, the method further includes: When a communication failure is detected in the Time-Sensitive Network (TSN) backbone, the central decision-making unit and each execution unit are switched to a traditional Ethernet-based communication mode, and a fixed-ratio power distribution strategy based on the rated power or rated load of each execution unit is implemented to maintain basic vehicle operational safety.

[0017] This invention has the following advantages: This invention designs a vehicle control architecture based on Time-Sensitive Network (TSN). The central decision-making unit synchronously collects the load, temperature, and health status of each local execution unit through TSN, and introduces dynamic constraints such as wheel-rail adhesion to perform multi-constraint adaptive optimization to achieve the global optimal allocation of traction and electric braking forces of each unit. Under braking conditions, a predictive-dynamic compensation mechanism is adopted to detect the decay of electric braking capacity in advance and smoothly introduce air braking compensation. The deterministic time window of TSN is used to achieve shock-free synchronous switching and closed-loop fine-tuning of dual braking commands, breaking the information silos of traditional networks and eliminating communication delay uncertainties. This achieves the global optimal allocation and precise collaborative control of the vehicle's power resources, effectively suppressing braking shock and improving ride comfort. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the core architecture of the system provided by the present invention. Detailed Implementation

[0019] Explanation of technical terms: TSN: Time-Sensitive Networking. A set of extended protocols based on standard Ethernet to provide deterministic latency, extremely low packet loss rate, and accurate time synchronization.

[0020] Deterministic latency refers to the time required for a data packet to travel from the source to the destination being predictable and bounded, without significant fluctuations due to changes in network load.

[0021] VCU / TCU / EBCU / BBCU: Vehicle Control Unit / Traction Control Unit / Electric Brake Control Unit / Air Brake Control Unit.

[0022] Jerk: Impact rate, which is the rate of change of deceleration (da / dt), measured in m / s³. It is a key indicator for measuring ride comfort; the lower the value, the more comfortable the ride.

[0023] FRER: Frame Replication and Elimination for Reliability, IEEE 802.1CB standard, improves reliability by sending redundant frames.

[0024] Global status awareness: refers to the central control unit's ability to acquire and understand the real-time operating information of all relevant subsystems of the vehicle, forming a unified and consistent operating view.

[0025] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] This invention designs a complete system integrating communication, control, and fault handling, the core architecture of which is as follows: Figure 1 The detailed descriptions of each part of the system are as follows: a) System hardware architecture and communication infrastructure Central Decision Unit (VCU): Employs a high-performance multi-core processor and integrates a power allocation algorithm module and a system health management module. It connects to the backbone network via a dual-port TSN network interface card.

[0027] TSN communication backbone: adopts a ring or dual-star topology, and the core switches support the key TSN standard set, including: IEEE 802.1AS-Rev (Time Synchronization): Achieves clock deviation of <1μs for all network devices.

[0028] IEEE 802.1Qbv (Time-Aware Shaper): Reserves a dedicated, periodic time window ("green channel") for the power control command stream, ensuring microsecond-level deterministic latency and zero packet loss.

[0029] IEEE 802.1CB (Frame Replication and Deletion): This feature allows for the simultaneous transmission of copies of critical command streams on redundant paths, ensuring that communication is not affected by any single point of failure.

[0030] Local execution units, including TCU, EBCU, and BBCU, must all be equipped with a TSN network interface and have precise clock synchronization capabilities and instruction execution timestamp feedback functions.

[0031] b) Multi-constraint adaptive dynamic allocation algorithm The algorithm runs in VCU, and its decision-making process is a constrained optimization problem.

[0032] Objective function: 1. Traction Condition: The objective is to achieve the highest overall vehicle traction efficiency or the lowest total energy consumption, min Σ(Loss_i(P_out)). Loss_i(P_out): The energy loss corresponding to the output power (P_out) of the i-th power unit. This loss model is mainly based on the experimental data of the traction system combination (i.e., the combined experiment of traction converter + motor) and vehicle energy consumption experimental data.

[0033] 2. Braking Condition: The primary objective is to maximize electric braking energy recovery, max Σ(P_eb_regen_i), where P_eb_regen_i is the energy recovery power of the i-th electric braking control unit (EBCU), while also considering braking comfort (minimizing the rate of change of deceleration). The objective function for braking condition is: max Σ(P_eb_regen_i) = λ·|da² / dt²|, where λ is the comfort weighting coefficient (calibrated through real vehicle testing, typically taken as 0.05~0.1), and da² / dt² is the second derivative of deceleration (i.e., the rate of change of Jerk).

[0034] Dynamic constraints: 1. Total Demand Constraint: Σ(F_traction_i) = F_total_demand (traction) or Σ(F_brake_i) = F_total_demand (braking). The sum of the traction forces of all power units must equal the total traction demand of the vehicle to ensure that the power requirements of the driver's operating commands (such as acceleration) are met. F_traction_i: The allocated traction force of the i-th unit; F_total_demand: The total traction demand of the vehicle (calculated from driver commands and signal system commands).

[0035] 2. Unit capacity constraint: F_min_i ≤ F_i ≤ F_max_i(T, Health). Here, F_max_i is not a fixed value, but is dynamically adjusted according to the motor / inverter temperature (T) and health status (Health). The Health index is mainly determined by fault information fed back from the traction system, where Health_i∈[0,1].

[0036] Maximum traction force / braking force: F_max_i = min[ (T_e_max_i · i_g) / r_wheel, μ_max · Load_i · g ].

[0037] Adhesion coefficient: μ_max = 0.35 - 0.001·v + 0.00001·v² (dynamically changes with vehicle speed v to adapt to wet / dry track conditions).

[0038] Minimum force constraint: F_min_i = -F_max_i / 2 (maximum allowable force for reverse braking / traction, to avoid component overload).

[0039] T_e_max_i: The maximum output torque of the motor in the i-th power unit (unit: N) m).

[0040] i_g: Transmission ratio (unitless), which is the torque transmission ratio from the motor output shaft to the wheel axle, used to amplify the motor torque and transmit it to the wheel.

[0041] r_wheel: Wheel rolling radius (unit: m), which refers to the distance from the point of contact between the wheel and the track to the center of the wheel.

[0042] μ_max: Maximum adhesion coefficient between wheel and rail (unitless), which changes dynamically with vehicle speed and the dryness or wetness of the rail, and represents the proportion of maximum friction force when wheel and rail do not slip.

[0043] Load_i: Real-time load (unit: kg) of the i-th carriage / power unit, obtained in real time from the air spring pressure signal, including the vehicle body weight + passenger / cargo weight. g: Gravitational acceleration (unit: m / s²), standard value 9.8 m / s², used to convert load (mass) into gravity (Load_i). g represents the total weight of the i-th unit (in N).

[0044] 3. Load Constraints: The load of each car is acquired in real time based on the pressure signal from the air spring, and this is used as one of the distribution benchmarks to optimize wheel-rail adhesion utilization. F_i = k_i · Load_i, where k_i = F_total_demand / ΣLoad_i (basic distribution coefficient), and k_i ∈ [k_min, k_max] (k_min / k_max are the upper and lower limits of the output ratio of each unit to avoid overloading of a single unit).

[0045] Meaning: The distributed force of the i-th unit is proportional to the real-time load of that unit, optimizing wheel-rail adhesion and preventing slippage. The real-time load of the i-th car (obtained from the air spring pressure signal, including the car body weight + passenger / cargo weight); 4. Cooperative Constraint (Dedicated to Braking): F_air_brake (Required Air Braking Force) = max(0, F_total_demand - Σ(F_eb_available_i)) ensures that the air braking force is non-negative (when the electric braking capacity is sufficient, the air brake does not intervene). Here, F_eb_available_i is the upper limit of the electric braking capacity calculated in real time.

[0046] The algorithm solves the above problem in each control cycle (e.g., 50ms) through a real-time optimizer and outputs a set of optimal force distribution instructions.

[0047] c) Predictive-Dynamic Compensation Electro-Pneumatic Co-braking Mechanism Prediction phase: The VCU continuously monitors the status of each EBCU and uses an extended Kalman filter (EKF) to predict the derating of electric braking capability. When it is predicted that an EBCU is about to be derated due to excessive temperature, it initiates cooperative braking preparation in advance (e.g., 500ms). Predictive model: (t+Δt) = f(x(t), u(t)) + w(t); State vector x(t) = [T_i(t), Health_i(t)], observation vector y(t) = [T_i_meas(t), I_eb_i_meas(t)]; Prediction step size Δt = 500ms, prediction accuracy: | (t+Δt) - y(t+Δt)|<5% The formula for the reduction in electric braking capacity is: F_eb_available_i(t+Δt) = F_eb_available_i(t) · e^(-k·ΔT), where: ΔT = T_i(t+Δt) - T_i(t), where k is the attenuation coefficient (calibrated by component characteristic test); F_eb_available_i(t+Δt): The available electric braking capacity (predicted value) of the i-th EBCU at time t+Δt; F_eb_available_i(t): The current available electric braking capacity of the i-th EBCU at time t (measured value); ΔT = T_i(t+Δt) - T_i(t): Predicted temperature change (ΔT>0 indicates temperature increase and reduced electric braking capacity); k: Attenuation coefficient (calibrated by component testing; different models of EBCU have different k values, reflecting the sensitivity of electric braking capability to temperature). e^(-k·ΔT): Exponential decay term (the higher the temperature and the larger k is, the more obvious the decay of electric braking capability, which is consistent with the physical characteristics of motor thermal decay).

[0048] Smooth transition phase: The command ramp is designed using an S-shaped transition curve (to avoid the step derivative of a linear ramp): Electric braking force command: F_eb_ramp(t) = F_eb_initial · [1 - 1 / (1 + e^(-a·(t -t0)))], t ∈ [t0, t0+T_trans] Air braking force command: F_air_ramp(t) = F_total_demand - F_eb_ramp(t) Parameter definition: a = 4 / T_trans (curve slope coefficient, controls transition smoothness), T_trans = 300~500ms (transition time, dynamically adjusted according to vehicle speed: the higher the vehicle speed, the longer T_trans). F_eb_ramp(t): Electric braking force command at time t (smoothly changing over time); F_eb_initial: Initial electric braking force (i.e., electric braking capability before decay) at the start time of switching (t0). t0: Start time of smooth switching (the moment when switching is initiated after the electric braking fades). t: Current time (valid only in the interval [t0, t0+T_trans], i.e., within the transition period); T_trans: Transition time (300~500ms, longer at higher vehicle speeds to ensure smoother switching at high speeds); a = 4 / T_trans: Curve slope coefficient (the longer T_trans is, the smaller a is, the gentler the S-curve, avoiding sudden changes in braking force); [1 - 1 / (1 + e^(-a·(t - t0)))]: S-shaped transition function (the derivative is continuous without abrupt changes, ensuring that the Jerk rate of deceleration is minimal).

[0049] TSN synchronization mechanism: Dedicated time windows are allocated through the Qbv gating list, and the time deviation of instruction issuance |t_issue_i - t_issue_j| < 1μs, ensuring synchronous response of the dual braking system.

[0050] Closed-loop fine-tuning stage: The VCU compares the command value with the actual output feedback from the EBCU / BBCU. If there is a slight deviation, dynamic compensation is immediately performed in the next cycle to readjust the output of other units and ensure that the total braking force remains constant. Deviation calculation: ΔF(t) = F_total_demand - [F_eb_actual(t) + F_air_actual(t)] Compensation strategy: ΔF_eb_i(t+T_ctrl) = ΔF(t) · (F_eb_cmd_i(t) / ΣF_eb_cmd_i(t)) · η_eb, ΔF_air_i(t+T_ctrl) = ΔF(t) · (F_air_cmd_i(t) / ΣF_air_cmd_i(t)) · η_air T_ctrl = 50ms (control cycle), η_eb / η_air is the braking efficiency correction coefficient (calibrated in real time by the feedback signal). Constraints: |ΔF_eb_i| ≤ 10%·F_eb_max_i, |ΔF_air_i| ≤ 15%·F_air_max_i (to avoid overcompensation leading to shocks).

[0051] in: F_eb_max_i: The maximum electric braking capacity of the i-th EBCU; 10%·F_eb_max_i: The upper limit of electric braking compensation in a single operation; F_air_max_i: The maximum air braking capacity of the i-th BBCU; 15%·F_air_max_i: The upper limit of single air braking compensation; |ΔF_eb_i| / |ΔF_air_i|: The absolute value of the compensation amount (ensuring it does not exceed the upper limit).

[0052] In special circumstances, the present invention can simplify the above solution as follows: 1. Algorithm Simplification Scheme When VCU computing resources are scarce, a fast allocation method based on fuzzy rules or a pre-defined MAP can be used as an alternative. Although this method is not as accurate as online optimization, it can still leverage the synchronization capabilities of the TSN to achieve better collaborative control than traditional methods. Specific details are as follows: (1) Fuzzy rule scheme: Input variables (all normalized to [0,1]): load deviation e_Load = (Load_i - Load_avg) / Load_max, temperature deviation e_T = (T_i - T_rated) / T_max, health status Health_i; Output variable: Allocation coefficient correction Δk_i ∈ [-0.2, 0.2] Example of core rule: ①If e_Load is large, e_T is small, and Health_i is good, then Δk_i is large; ②If e_Load is small, e_T is large, and Health_i is poor, then Δk_i is small; Reasoning method: Mamdani method; defuzzification: centroid method.

[0053] (2) Preset MAP diagram scheme: MAP plot dimensions: 2D (Load_i × Temperature T_i) → Allocation coefficient k_i, 3D (Load_i × T_i × Health_i) → Correction coefficient k_corr; Interpolation algorithm: Bilinear interpolation k_i = a·k1 + b·k2 + c·k3 + d·k4 (a+b+c+d=1, calculated from 4 adjacent MAP points); The algorithm simplifies storage and updates: The MAP graph is stored in the VCU flash memory and supports remote updates via the TSN network, with an update cycle of ≤10 minutes.

[0054] 2. Network downgrade solution: When the TSN network fails completely, the system can switch to a best-effort communication mode based on traditional Ethernet and implement a simplified, fixed-ratio power distribution strategy to maintain basic vehicle operational safety, as follows: (1) Fixed proportion allocation strategy: ① Traction condition: F_i = F_total_demand · (P_rated_i / ΣP_rated_i) (allocated according to the rated power of the unit); ② Braking condition: F_eb_i = F_eb_total · (Load_rated_i / ΣLoad_rated_i), F_air_i = F_air_total · (Load_rated_i / ΣLoad_rated_i) (distributed according to rated load).

[0055] (2) Communication delay compensation: t_issue = t_cmd - Δt_comp, where Δt_comp is the average delay of traditional Ethernet (pre-calibrated to 20~50ms).

[0056] (3) Fault switching logic: The TSN network failure judgment condition is "no synchronization clock signal received for 3 consecutive control cycles", the switching time is ≤100ms, and the braking force fluctuation during the switching process is ≤±8%.

[0057] Compared with existing technologies, the benefits of this invention are significant and multifaceted: thanks to the nanosecond-level synchronization and global algorithm optimization of TSN, the control synchronization error is reduced from the millisecond level to the microsecond level, a reduction of more than 85%, and the longitudinal impact rate during electro-pneumatic braking switching is reduced by more than 50%, significantly improving the subjective evaluation of passenger comfort; adaptive electric braking load distribution improves energy recovery efficiency by 5% to 10%, and equipment load balancing leads to an expected 15% extension of mean time between failures, effectively reducing the total life cycle cost; dual-network redundancy and the TSN fault self-healing mechanism improve system availability to over 99.999%, laying a safe foundation for the networking of core control functions; at the same time, this invention proactively realizes the prototype of a "software-defined vehicle power" architecture, which can flexibly upgrade the control algorithm or expand functions through software updates, greatly enhancing the system's adaptability to future technology iterations.

[0058] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of the present invention are not limited to the described order of actions, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to the embodiments of the present invention.

[0059] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0060] The various embodiments in this specification are described in a related manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0061] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.

Claims

1. A vehicle power adaptive allocation and cooperative control method based on time-sensitive networks, characterized in that, The method includes: The central decision-making unit is connected to multiple execution units through a communication backbone based on Time Sensitive Network (TSN), and the clocks between the central decision-making unit and each execution unit are synchronized using the TSN time synchronization protocol. The central decision-making unit periodically obtains the total power demand command of the vehicle and the real-time status information fed back by each execution unit. The real-time status information includes at least the load information of the compartment corresponding to each execution unit, the temperature information of each execution unit, and the health status information. Based on the received real-time status information, the central decision-making unit determines the dynamic capability constraints of each execution unit under the current operating conditions. Under the premise of satisfying the total power demand constraints of the vehicle, the dynamic capability constraints of each execution unit, and the load constraints based on load information, the central decision-making unit calculates the power distribution command of each execution unit through a preset optimization allocation algorithm. In the braking condition, it also includes a cooperative constraint so that the required air braking force is determined based on the comparison between the total braking demand and the total available electric braking capacity of each electric braking unit. The central decision-making unit adopts the time-aware shaping mechanism of the TSN network to reserve a transmission time window with deterministic delay for power allocation commands, and synchronously sends each power allocation command to the corresponding execution unit within the same control cycle; Each execution unit performs traction or braking actions according to the received power distribution command and feeds back the execution status, including the actual output value, to the central decision-making unit.

2. The method according to claim 1, characterized in that, The TSN time synchronization protocol is used to synchronize the clocks between the central decision-making unit and each execution unit, including: The IEEE 802.1AS protocol is used to synchronize the clocks between the central decision-making unit and each execution unit.

3. The method according to claim 1, characterized in that, Load information is obtained in real time by collecting the pressure signals of the air springs in each carriage; temperature information is obtained in real time by collecting the temperature sensors of the motors or inverters in each execution unit; and health status information is determined comprehensively based on the fault diagnosis information fed back by each execution unit.

4. The method according to claim 1, characterized in that, The dynamic capability constraints of each execution unit under the current operating conditions include: The maximum traction or braking force of each actuator under the current operating conditions is limited by the temperature and health status information of the actuator, and also by the wheel-rail adhesion conditions determined based on the real-time vehicle speed and wheel-rail condition. The minimum traction force or minimum braking force of each actuator is limited by the reverse allowable ratio of the maximum traction force or maximum braking force of each actuator under the current operating conditions.

5. The method according to claim 1, characterized in that, The preset optimized allocation algorithm: Under traction conditions, the traction force of each actuator is distributed with the goal of minimizing the total energy consumption of the vehicle. Under braking conditions, the distribution of electric braking force and air braking force among the actuators is prioritized with maximizing the energy recovery of electric braking while also taking into account braking comfort.

6. The method according to claim 5, characterized in that, The preset optimization allocation algorithm, under braking conditions, takes maximizing the sum of available electric braking capabilities provided by each electric braking control unit as the optimization objective, and introduces a comfort weight coefficient calibrated through a real vehicle to constrain the deceleration rate of change, so as to adjust the allocation relationship between electric braking and air braking.

7. The method according to claim 1, characterized in that, The method further includes: When the central decision-making unit has limited computing resources, the optimized allocation algorithm is replaced with a fast allocation method based on fuzzy rules or a preset MAP. The fuzzy rules take the load deviation, temperature deviation and health status of each execution unit as input variables and output the correction amount of the allocation coefficient. The preset MAP stores the allocation coefficient with load and temperature as dimensions and calculates the real-time allocation coefficient through an interpolation algorithm.

8. The method according to claim 1, characterized in that, The method further includes: During braking, the central decision-making unit continuously monitors the real-time status of each electric braking control unit. When it is predicted that the electric braking capacity of any electric braking control unit will decrease due to temperature rise, a transition command including gradual reduction of electric braking force and gradual increase of air braking force is generated in advance to control the smooth switching of the electric braking system and the air braking system. The electric braking force command and the air braking force command in the transition command are generated using an S-shaped transition curve with continuous derivatives to ensure that the total braking force remains continuous throughout the transition period, while the deceleration rate of change is maintained within the preset comfort range. After a smooth transition, the central decision-making unit compares the deviation between the actual output force fed back by each electric braking control unit and the corresponding command value with the actual output force fed back by each air braking control unit. In the next control cycle, it dynamically compensates and adjusts the power distribution command of each execution unit according to the deviation to ensure the stability of the total braking force of the vehicle.

9. The method according to claim 8, characterized in that, The method further includes: An extended Kalman filter algorithm is used to predict the temperature change trend after a preset time by using the temperature and health status of the electric braking control unit as state variables, and to calculate the corresponding predicted value of the available electric braking capacity based on the attenuation law calibrated by the component characteristics.

10. The method according to claim 1, characterized in that, The method further includes: When a communication failure is detected in the Time-Sensitive Network (TSN) backbone, the central decision-making unit and each execution unit are switched to a traditional Ethernet-based communication mode, and a fixed-ratio power distribution strategy based on the rated power or rated load of each execution unit is implemented to maintain basic vehicle operational safety.

Images