Bridge pier active anti-collision system control method and system based on fuzzy neural network

Abstract

The invention belongs to the technical field of intelligent control, and particularly relates to a fuzzy neural network-based bridge pier active anti-collision system control method and system, and the method comprises the steps: obtaining the identity attribute data and motion state data of an incoming ship in a target water area; determining the design load and the drainage volume of the ship according to the identity attribute data, and calculating the equivalent impact kinetic energy of the ship in combination with the speed to the ground and the impact incident angle; a fuzzy neural network is constructed, the equivalent impact kinetic energy and the distance between the ship and the anti-collision facility serve as input variables, and basic damping current for controlling the magneto-rheological damper is output; a temperature change compensation correction coefficient is determined according to the difference value between the real-time temperature of the cylinder body and the calibration reference temperature, and final self-adaptive control current is generated through the temperature change compensation correction coefficient so as to drive the magnetorheological damper. The magnetorheological damper can adapt to collision of ships of different tonnages, and performance attenuation caused by temperature rise of the magnetorheological damper is effectively restrained.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology. More specifically, this invention relates to a control method and system for an active collision avoidance system for bridge piers based on fuzzy neural networks. Background Technology

[0002] Bridge pier anti-collision facilities are an important barrier to ensure the safety of bridge structures. Their development has evolved from rigid anti-collision devices to flexible anti-collision devices. Most of the current bridge pier anti-collision facilities are passive structures, and their stiffness and damping characteristics are fixed after they are manufactured. When facing collisions from ships of different tonnages, they are difficult to achieve the same protective effect.

[0003] In existing technologies, bridge pier collision protection systems with fixed stiffness and damping characteristics face a dilemma in accommodating vessels of different tonnages: if the bridge pier collision protection device is designed too rigidly, the excessive reaction force in the event of a collision with a small tonnage vessel can cause damage or even complete destruction to the vessel; if the bridge pier collision protection device is designed too softly, a collision with a large tonnage vessel can easily penetrate the bridge pier collision protection device, directly damaging the bridge pier. Furthermore, passive bridge pier collision protection systems often generate enormous peak reaction forces at the moment of high-speed ship impact, and this instantaneous impact force can cause significant damage to the bridge pier foundation and the ship's hull structure.

[0004] There are existing theories proposing the use of magnetorheological dampers. However, in actual impacts, the enormous mechanical energy is rapidly converted into heat energy, causing the temperature of the magnetorheological damper to rise sharply. The physical properties of the magnetorheological fluid in the damper determine that its viscosity will decrease significantly with increasing temperature, resulting in a decrease in the output force of the damper and causing the damper to fail. Summary of the Invention

[0005] To address the problems of fixed stiffness and poor adaptability of existing bridge pier collision protection facilities, as well as the thermal attenuation effect of magnetorheological dampers, this invention provides a control method and system for an active bridge pier collision protection system based on a fuzzy neural network. This system can adapt to collisions with ships of different tonnages and effectively suppress the performance degradation of magnetorheological dampers caused by temperature rise, thereby improving the reliability and protective effect of the system.

[0006] In a first aspect, the present invention provides a control method for an active anti-collision system for bridge piers based on a fuzzy neural network, comprising: acquiring the identity attribute data and motion state data of an approaching vessel in a target water area, and smoothing the motion state data to eliminate clutter interference; determining the design load and displacement volume of the vessel based on the identity attribute data, and calculating the equivalent impact kinetic energy of the vessel by combining the ground speed and the impact incident angle relative to the anti-collision surface in the motion state data; constructing a fuzzy neural network, taking the equivalent impact kinetic energy and the distance between the vessel and the anti-collision facility as input quantities, and outputting the basic damping current for controlling the magnetorheological damper through fuzzification, rule reasoning and defuzzification processes; acquiring the real-time temperature of the cylinder of the magnetorheological damper, determining the temperature change compensation correction coefficient based on the difference between the real-time cylinder temperature and the calibrated reference temperature, correcting the basic damping current using the temperature change compensation correction coefficient, and superimposing a feedback term based on the relative compression speed of the damper piston to generate a final adaptive control current to drive the magnetorheological damper.

[0007] By adopting the above technical solution, the system can output a nonlinear basic damping current using a fuzzy neural network based on the actual impact kinetic energy and distance of the ship. Furthermore, it addresses the performance degradation problem of the magnetorheological damper during impact heating through a temperature-varying compensation correction coefficient. This control method enables the bridge pier anti-collision device to provide flexible energy absorption to protect the hull when a small-tonnage ship impacts the bridge pier, and to provide rigid blocking to protect the bridge pier when a large-tonnage ship impacts it. This effectively solves the problem of traditional passive anti-collision facilities having fixed stiffness and being unable to meet the protection needs of ships of varying tonnages.

[0008] Preferably, the acquisition of the identity attribute data and motion status data of the approaching vessel in the target waters includes: parsing the unique identification code of the approaching vessel through the Automatic Identification System (AIS) to obtain the design load, length, beam, and draft as identity attribute data; tracking the approaching vessel using millimeter-wave radar and lidar to obtain the ground speed, heading angle, and incident angle relative to the collision avoidance surface as motion status data; and processing the raw data collected by the millimeter-wave radar and lidar using a Kalman filter algorithm.

[0009] By employing the above technical solution, and acquiring the identity attribute data and motion status data of approaching vessels, comprehensive and accurate perception of the identity and motion status of approaching vessels within the target waters is achieved. By using a Kalman filter algorithm to smooth the raw motion status data acquired by the radar, clutter interference from water waves and environmental noise can be effectively eliminated, and outliers such as speed jumps can be removed. This provides high-confidence input data that conforms to the laws of physical motion for subsequent kinetic energy calculations and control decisions.

[0010] Preferably, the equivalent impact kinetic energy satisfies the following relationship: ; In the formula, For equivalent impact kinetic energy, For the design load capacity of the ship, For the coefficient of the attached water, For water density, For drainage volume, For the ship's speed over land, The angle of impact incidence.

[0011] By adopting the above technical solution and using an equivalent impact kinetic energy calculation formula that includes the attached water coefficient and the impact incident angle, compared with the traditional kinetic energy formula that only considers the mass of the ship itself, this solution fully takes into account the mass of the water moving with the hull and the influence of the impact angle on the effective energy. It avoids the impact kinetic energy estimation deviation caused by ignoring the inertia of the water or failing to strip away the ineffective tangential energy, ensuring that the collision avoidance system can respond based on the real destructive energy, and significantly improving the scientificity and accuracy of impact energy assessment.

[0012] Preferably, the construction of the fuzzy neural network includes: setting an input layer to receive the equivalent impact kinetic energy and the distance; setting a fuzzification layer to convert the input quantity into the membership degree of the corresponding fuzzy linguistic variable using a membership function; setting a rule inference layer to perform logical operations on the membership degree according to a preset fuzzy control rule library to activate the corresponding control rule; and setting an output layer to perform defuzzification calculation on the output result of the rule inference layer to obtain the basic damping current.

[0013] By adopting the above technical solution, a nonlinear mapping model between energy, distance, and damping current is constructed using a fuzzy neural network. By setting up an input layer, a fuzzification layer, a rule-based reasoning layer, and an output layer, the system can handle fuzzy linguistic variables such as high energy and close-range conditions, and automatically derive the optimal basic damping current based on preset logic. This enables the collision avoidance system to smoothly handle complex impact conditions and achieve adaptive adjustment of the damping force.

[0014] Preferably, the fuzzy control rule base includes establishing a nonlinear mapping relationship between energy, distance and damping current, and when the equivalent impact kinetic energy is large and the distance is small, an increased basic damping current is output.

[0015] Preferably, the temperature change compensation correction coefficient satisfies the following relationship: ; in, This is the temperature change compensation correction factor. This refers to the real-time temperature of the damper cylinder. To calibrate the reference temperature, It is the thermal viscosity attenuation factor. It is an exponential factor.

[0016] By employing the above technical solution, the cylinder temperature of the magnetorheological damper is monitored in real time, and a correction coefficient is calculated using a formula that includes a thermal viscosity decay factor and an exponential factor. When the temperature rises and the viscosity of the magnetorheological fluid decreases, the control current is automatically increased to compensate for the loss of damping force. This ensures that even in the event of continuous impacts or high-energy impacts that cause a large amount of mechanical energy to be converted into heat energy, the collision avoidance system can still maintain the expected protective performance, avoiding control failure due to overheating.

[0017] Preferably, the adaptive control current satisfies the following relationship: ; In the formula, For the final adaptive control current, Based on the basic damping current, The relative compression velocity of the damper piston. For speed feedback coefficient, For the amplitude limiting function, This is the upper limit for safe current.

[0018] Preferably, the limiting function is used to limit the calculated current value between 0 and the current safety upper limit value to prevent reverse current or overload current.

[0019] Preferably, the magnetorheological damper is filled with a magnetorheological fluid, the viscosity of which decreases as the temperature increases, and the temperature change compensation correction coefficient is used to increase the current output to compensate for the viscosity loss when the temperature increases.

[0020] Secondly, the present invention provides an active anti-collision system for bridge piers based on fuzzy neural networks, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned control method for the active anti-collision system for bridge piers based on fuzzy neural networks is implemented.

[0021] By adopting the above technical solution, the above-mentioned control method for the active anti-collision system of bridge piers based on fuzzy neural networks is generated into a computer program and stored in a memory so that it can be loaded and executed by a processor. In this way, a terminal device can be made based on the memory and the processor for convenient use.

[0022] The beneficial effects of this invention are as follows: This invention proposes a closed-loop control architecture for perception-decision compensation execution, which utilizes a fuzzy neural network as the brain to process nonlinear relationships during the impact process and achieve macroscopic stiff-flexible switching. Furthermore, this invention creatively introduces a thermodynamic-based temperature change compensation mechanism. By monitoring the damper temperature in real time and dynamically adjusting the current, it solves the industry problem of the failure of the protective force of the magnetorheological anti-collision system due to the thinning of the working fluid under continuous impact or high temperature environment. Combined with the attached water quality correction in physical modeling, it not only ensures the intelligence of the control strategy, but also ensures the physical safety and performance stability of the system under all weather and all operating conditions. Attached Figure Description

[0023] Figure 1 This is a flowchart of the control method for the bridge pier active anti-collision system based on fuzzy neural network in an embodiment of the present invention; Figure 2 These are adaptive control current response curves under different impact energy levels in embodiments of the present invention; Figure 3 This is a time-history comparison diagram of impact force between active control and passive protection in an embodiment of the present invention. Detailed Implementation

[0024] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0025] This invention discloses a control method for an active collision avoidance system for bridge piers based on a fuzzy neural network, referring to... Figure 1 This includes steps S1 to S4: S1. Acquire the identity attribute data and motion status data of incoming ships in the target water area, and smooth the motion status data to eliminate clutter interference.

[0026] In an optional embodiment, the present invention acquires the identity attribute data and motion status data of approaching vessels in the target waters by establishing a multi-source situational awareness mechanism. Specifically, the system receives and parses the unique identification code of the approaching vessel through an Automatic Identification System (AIS), thereby retrieving the vessel's information, including its design deadweight tonnage, from the database. Static identity attribute data such as ship length, beam, and draft.

[0027] While acquiring the static identity attribute data of the approaching vessel, the system uses millimeter-wave radar and / or lidar deployed around the bridge piers to scan the water surface in real time, tracking the dynamic trajectory of the approaching vessel and acquiring its raw motion state data. The raw motion state data acquired by the system using the detection equipment includes the vessel's ground speed. Information such as heading angle and impact incident angle relative to the surface of the anti-collision device is also included. Since water waves and environmental noise may introduce high-frequency clutter into the radar data, in this embodiment of the invention, the raw motion state data obtained by the detection device also needs to be smoothed using the Kalman filter algorithm. For example, when the speed data detected by the detection device undergoes a non-physical drastic change in a short period of time, the Kalman filter algorithm will remove outliers based on the ship's motion inertia model and retain only the smooth trajectory data.

[0028] In this way, by fusing and cleaning multi-source data, accurate physical input can be provided for subsequent decision-making, avoiding misjudgments caused by clutter interference and ensuring the reliability of system perception.

[0029] S2. Determine the ship's design load and displacement volume based on the identity attribute data, and calculate the ship's equivalent impact kinetic energy by combining the ground speed and impact incident angle relative to the anti-collision surface in the motion state data.

[0030] In an alternative embodiment, in water collision avoidance engineering, if the kinetic energy formula is simply applied... This introduces significant errors, specifically: because the ship navigates in water, the water surrounding the hull also moves with the ship. If the mass of the water moving with the hull is not considered, the system will severely underestimate the impact energy. Since the normal component of the ship's impact velocity causes crushing damage, and the tangential component mainly generates friction, the ineffective tangential energy must be stripped away. Therefore, this invention calculates the equivalent impact kinetic energy using the following method. : ; in, This is the equivalent impact kinetic energy; For the design load capacity of the ship; The coefficient for adhering water is set to 0.05-0.1 in this invention; Density of water; This refers to the drainage volume; This refers to the ship's speed over land. The angle of impact incidence.

[0031] To more clearly illustrate the equivalent impact kinetic energy The function and calculation process will be explained with examples below: First, let's assume the design load of the ship colliding with the bridge pier is... kg; drainage volume is 2000 cubic meters; the adjoining water coefficient is 0.05; water density is The ship's speed above ground is The impact incident angle is Substituting this into the formula for the equivalent impact kinetic energy, we can obtain... The value is 4728522. However, if the influence of the attached water mass and the correction angle is not considered, the calculated impact kinetic energy will have a large deviation. This formula ensures the scientific nature of the energy estimation.

[0032] Thus, by introducing the mass and angle correction of the attached water, the actual destructive energy carried by the ship in the water can be accurately assessed, avoiding energy estimation errors caused by ignoring the ineffective tangential components of water inertia and speed, and providing a real physical basis for the control system.

[0033] S3. Construct a fuzzy neural network, taking the equivalent impact kinetic energy and the distance between the ship and the collision avoidance facility as input quantities. Through fuzzification, rule reasoning and defuzzification processes, output the basic damping current that controls the magnetorheological damper.

[0034] In an optional embodiment, since the impact of a ship is a nonlinear process, when the current driving the magnetorheological damper is a constant or linearly changing value, it cannot match the impact process of the ship in real time. Therefore, the present invention uses the nonlinear mapping capability of the neural network to achieve intelligent stiffness adjustment.

[0035] The neural network used in this invention has a multi-layered structure: an input layer, a fuzzification layer, a rule-based reasoning layer, and an output layer. The input layer is used to receive physical quantities. as well as The fuzzification layer uses a Gaussian function to convert specific numerical values ​​into membership degrees of low, medium, and high fuzzy linguistic variables. For example, an energy of 4.7 MJ may correspond to high energy, with a membership degree of 0.8. The rule inference layer performs fuzzy logic AND operations, activating preset expert rules. The system internally stores multiple rules; for example, if the energy is extremely high and the distance is extremely short, maximum stiffness must be activated. Finally, the output layer performs defuzzification and outputs a basic damping current. This is a prediction-based open-loop instruction, which does not yet take into account the state of the actuator.

[0036] Thus, by leveraging the nonlinear mapping capability of fuzzy neural networks, intelligent stiffness adjustment can be achieved, adapting to stronger or weaker conditions, thus solving the problem that traditional anti-collision devices, due to their fixed stiffness, cannot accommodate ships of varying sizes.

[0037] S4. Collect the real-time temperature of the cylinder of the magnetorheological damper, determine the temperature change compensation correction coefficient based on the difference between the real-time temperature of the cylinder and the calibrated reference temperature, use the temperature change compensation correction coefficient to correct the basic damping current, and superimpose the feedback term based on the relative compression speed of the damper piston to generate the final adaptive control current to drive the magnetorheological damper.

[0038] In an optional embodiment, the magnetorheological damper converts kinetic energy into heat energy during severe impacts in practical applications. This heat causes the magnetorheological fluid to thin out as the damper's temperature rises. If the original output current cannot achieve the preset damping force, this invention introduces a temperature-change compensation correction coefficient. To compensate for the command current. The calculation method is as follows: ; in, This is the temperature change compensation correction factor; The real-time temperature of the damper cylinder is obtained by a temperature sensor installed on the magnetorheological damper. For calibrating the reference temperature; It is the thermal viscosity attenuation factor; It is an exponential factor.

[0039] To more clearly illustrate the role and calculation process of the temperature change compensation correction factor, the following example will demonstrate this: First, set the calibration reference temperature. At 25 degrees Celsius, the thermal viscosity attenuation factor It is 0.005; the index factor. The value is 1; when continuous impacts cause the cylinder block temperature to rise... Rise to hour, This means that the system has detected the temperature rise of the magnetorheological damper and has automatically amplified the command current by 1.15 times to compensate for the performance loss caused by the thinning of the magnetorheological fluid.

[0040] After obtaining the temperature change compensation correction coefficient, the system calculates the final adaptive control current by combining the speed feedback: ; in, This is the final current applied to the coil; This is the temperature change compensation correction factor; This is a temperature-compensated feedforward term; For feedback items; The relative compression velocity of the piston in the magnetorheological damper increases with increasing impact speed. The larger the current, the more it automatically increases; when the impact ends and the ship retracts... When the current becomes negative, it automatically decreases, and the auxiliary device resets. The function ensures that the current is limited to Within the safe zone.

[0041] Reference Figure 2The system current response exhibits an S-shaped curve, maintaining extremely low current under low-energy impacts to achieve flexible energy absorption, linearly increasing current during intermediate-energy impacts, and saturating under high-energy impacts, thus transitioning to rigid protection. (Refer to...) Figure 3 The active control curve of this invention is a platform-shaped trapezoid with a peak value significantly lower than that of the existing passive protection curve and a longer action time, resulting in excellent energy management performance.

[0042] Thus, by introducing a temperature change compensation mechanism and velocity feedback, the system effectively suppresses the thermal decay effect of the magnetorheological fluid, ensuring the agility and stability of the action response across the entire temperature range, thereby integrating the accuracy of physical modeling with the adaptability of data-driven approaches.

[0043] This invention also discloses an active anti-collision system for bridge piers based on a fuzzy neural network, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the control method for the active anti-collision system for bridge piers based on a fuzzy neural network according to this invention is implemented.

[0044] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0045] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise expressly and specifically defined.

Claims

1. A control method for an active collision avoidance system for bridge piers based on fuzzy neural networks, characterized in that, include: Acquire the identity attribute data and motion status data of incoming ships in the target water area, and smooth the motion status data to eliminate clutter interference; The design load and displacement volume of the ship are determined based on the identity attribute data. The equivalent impact kinetic energy of the ship is calculated by combining the ground speed and the impact incident angle relative to the anti-collision surface in the motion state data. A fuzzy neural network is constructed, taking the equivalent impact kinetic energy and the distance between the ship and the collision avoidance facility as inputs. Through fuzzification, rule reasoning and defuzzification processes, the basic damping current controlling the magnetorheological damper is output. The real-time temperature of the cylinder of the magnetorheological damper is collected. The temperature change compensation correction coefficient is determined based on the difference between the real-time temperature of the cylinder and the calibrated reference temperature. The basic damping current is corrected using the temperature change compensation correction coefficient. A feedback term based on the relative compression speed of the damper piston is superimposed to generate the final adaptive control current to drive the magnetorheological damper.

2. The control method for the bridge pier active collision avoidance system based on fuzzy neural network according to claim 1, characterized in that, The acquisition of the identity attribute data and movement status data of incoming vessels within the target waters includes: The unique identification code of the incoming vessel is analyzed by the Automatic Identification System (AIS) to obtain the design load capacity, length, beam and draft as identity attribute data. Millimeter-wave radar and lidar are used to track oncoming ships and obtain their speed to the ground, heading angle, and incident angle relative to the collision avoidance surface as motion status data. The raw data acquired by the millimeter-wave radar and lidar are processed using the Kalman filter algorithm.

3. The control method for the bridge pier active collision avoidance system based on fuzzy neural network according to claim 1, characterized in that, The equivalent impact kinetic energy satisfies the following relationship: ； In the formula, For equivalent impact kinetic energy, For the design load capacity of the ship, For the coefficient of the attached water, For water density, For drainage volume, For the ship's speed over land, The angle of impact incidence.

4. The control method for the bridge pier active collision avoidance system based on fuzzy neural network according to claim 1, characterized in that, The construction of the fuzzy neural network includes: The input layer is configured to receive the equivalent impact kinetic energy and the distance; Set up a fuzzification layer and use a membership function to convert the input quantity into the membership degree of the corresponding fuzzy linguistic variable; A rule reasoning layer is set up to perform logical operations on the membership degree according to a preset fuzzy control rule library, thereby activating the corresponding control rules. An output layer is set up, and the output results of the rule reasoning layer are defuzzified to obtain the basic damping current.

5. The control method for the bridge pier active collision avoidance system based on fuzzy neural network according to claim 4, characterized in that, The fuzzy control rule base includes establishing a nonlinear mapping relationship between energy, distance and damping current. When the equivalent impact kinetic energy is large and the distance is small, the basic damping current is increased.

6. The control method for the bridge pier active collision avoidance system based on fuzzy neural network according to claim 1, characterized in that, The temperature change compensation correction coefficient satisfies the following relationship: ； in, This is the temperature change compensation correction factor. This refers to the real-time temperature of the damper cylinder. To calibrate the reference temperature, It is the thermal viscosity attenuation factor. It is an exponential factor.

7. The control method for the bridge pier active collision avoidance system based on fuzzy neural network according to claim 6, characterized in that, The adaptive control current satisfies the following relationship: ； In the formula, For the final adaptive control current, Based on the basic damping current, The relative compression velocity of the damper piston. For speed feedback coefficient, For the amplitude limiting function, This is the upper limit for safe current.

8. The control method for the bridge pier active collision avoidance system based on fuzzy neural network according to claim 7, characterized in that, The limiting function is used to limit the calculated current value between 0 and the safe upper limit of the current to prevent reverse current or overload current.

9. The control method for the bridge pier active collision avoidance system based on fuzzy neural network according to claim 1, characterized in that, The magnetorheological damper is filled with magnetorheological fluid, the viscosity of which decreases as the temperature increases. The temperature-change compensation correction coefficient is used to increase the current output to compensate for the viscosity loss when the temperature increases.

10. A bridge pier active collision avoidance system based on fuzzy neural network, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the control method for an active anti-collision system for bridge piers based on a fuzzy neural network according to any one of claims 1-9.

Images