Wind power tower lifting platform based on adaptive weight and fuzzy PID control and adjusting method thereof

Abstract

The invention discloses a wind power tower lifting platform based on self-adaptive weight and fuzzy PID control and an adjusting method of the wind power tower lifting platform. The lifting platform comprises a main frame body, a ladder stand component, a guide rail assembly, a lifting platform mechanism and a lifting driving assembly. The device comprises a transmission chain wheel assembly, a driven chain wheel assembly, a counterweight assembly, an anti-falling assembly, a top corridor, a sliding terminal and a conductive sliding frame. The adjusting method comprises the steps of sensor installation and data acquisition; processing and analyzing data; self-adaptive weight optimization adjustment is carried out, the load condition and the environment condition of the current platform are judged, and the optimal balancing weight position is calculated in real time; ensuring that the lifting speed is matched with the current load according to a weight optimization result; and carrying out multi-layer neural network load prediction. The stability and safety of the platform under complex working conditions are improved, and the problems that a traditional method is slow in response and insufficient in adjusting precision when dealing with load fluctuation and environment change are solved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control and automation, and in particular to a wind power tower lifting platform based on adaptive weight and fuzzy PID control and an adjustment method thereof. Background Art

[0002] In the field of offshore wind power operation and maintenance, the balance adjustment of offshore wind power tower lifting platforms is of great significance. Traditional balance adjustment methods usually rely on manual experience and fixed counterweight adjustment. However, this method has many limitations and defects. For example, the response speed is slow, the accuracy is not high, and it is impossible to make timely adjustments under complex working conditions, which easily leads to platform imbalance.

[0003] In recent years, with the development of automation and intelligent control technology, adaptive control and PID control, as an efficient automatic adjustment method, make up for the shortcomings of traditional manual adjustment methods by real-time monitoring and dynamic adjustment of platform loads. However, the existing automatic control system still has many problems in dealing with complex and changeable working conditions. For example, how to deal with drastic fluctuations in loads and changes in environmental conditions is still a challenge facing system stability.

[0004] Traditional PID control algorithms have certain limitations. For example, they cannot quickly adapt to large load changes and environmental fluctuations. To this end, researchers have proposed an algorithm based on adaptive weight optimization and fuzzy PID control, which improves the platform's dynamic response capability and adjustment accuracy by real-time monitoring of load and environmental parameters. However, in practical applications, how to effectively combine these improved algorithms to form a systematic balance adjustment method for offshore wind power tower lifting platforms still requires further research and exploration. Summary of the invention

[0005] Purpose of the invention: The purpose of the present invention is to provide a wind turbine tower lifting platform and an adjustment method thereof based on adaptive weight and fuzzy PID control.

[0006] Technical solution: The wind power tower lifting platform based on adaptive weight and fuzzy PID control described in the present invention includes a main frame, a ladder component, a guide rail assembly, a lifting platform mechanism, a lifting drive assembly, a transmission sprocket assembly, a driven sprocket assembly, a counterweight assembly, an anti-fall assembly, a top corridor, a sliding terminal and a conductive slide. The main frame is the supporting structure of the lifting platform and is installed on the outside of the offshore wind power tower body. The ladder component is installed on the main frame, and the guide rail assembly is arranged along the main frame. The lifting platform mechanism is arranged on the main frame and cooperates with the guide rail assembly and the ladder component to move up and down along the main frame for lifting people or objects. A transmission sprocket assembly and a driven sprocket assembly are arranged at the top of the ladder component. The lifting platform mechanism is connected to the counterweight assembly, and the lifting and lowering are controlled by increasing or decreasing the weight of the counterweight assembly. The ladder component is also provided with an anti-fall assembly.

[0007] Furthermore, the lifting platform mechanism comprises a connecting head, a platform body, a platform frame, and a slider, and the lower end of the guide rail assembly is provided with a connecting head to connect the lifting platform mechanism and the chain, and the lower end of the platform body is provided with a platform frame (43), and the platform frame and the slider are connected to move on the guide rail assembly.

[0008] Furthermore, a pull-down protective rod, a guard plate and a rotating anti-fall sleeve are arranged on the side of the platform body, and a top corridor and a conductive slide are arranged on the other side. A sliding terminal is arranged on the conductive slide, and a protective strip is arranged between the conductive slide and the sliding terminal. A hanging buckle and a flip stack are arranged on the top of the platform body.

[0009] The wind power tower lifting platform adjustment method based on adaptive weight and fuzzy PID control of the present invention comprises the following steps:

[0010] S1. Sensor installation and data collection, including platform load, wind speed, platform motion status and other data, to ensure that the collected data can fully reflect the operating status of the platform;

[0011] S2, data processing and analysis;

[0012] S3, adaptive weight optimization adjustment, judges the current platform load and environmental conditions based on the data of S2, and calculates the optimal counterweight position in real time;

[0013] S4, application of fuzzy PID control algorithm, to ensure that the lifting speed matches the current load according to the weight optimization result of S3;

[0014] S5, multi-layer neural network load prediction, predicts load changes in advance by analyzing the real-time lifting status and load information in S4;

[0015] S6. Security monitoring and feedback mechanism to ensure that abnormal situations in all links can be captured in a timely manner.

[0016] Furthermore, step S1 includes: collecting the load data, wind speed data and motion state data of the platform, and transmitting all the collected motion state data to the central control system through the data interface. The system analyzes these data in real time and combines the load data and wind speed data to ensure the stability of the platform operation. If the system detects an abnormality in the motion state, the control system will automatically adjust the operating speed of the platform.

[0017] Furthermore, the load data of the platform is collected by installing strain gauge load sensors at four key supporting points at the bottom of the platform. These sensors monitor the load on the platform in real time and convert mechanical force into electrical signals through precise strain detection and transmit them to the central control system through amplifiers. The signals are converted into digital signals through analog-to-digital converters and input into the control system for data processing.

[0018] The real-time total load Q of the platform is determined by the output signal F of each sensor. i The calculation formula is:

[0019]

[0020] Where Q is the total load of the platform, F i is the load value detected by the i-th sensor, and n is the number of sensors installed on the platform.

[0021] Furthermore, the acquisition of the motion state data of the platform includes installing a motion sensor on the platform frame, near the platform rail, detecting the vertical position x(t) of the platform in real time, and calculating the displacement by integrating the velocity v(t):

[0022] x(t)=x0+∫0 t v(t)dt

[0023] Among them, x0 represents the initial position of the platform, v(t) represents the instantaneous velocity of the platform, which is detected by the velocity sensor installed on the drive system or guide rail assembly. At the same time, an acceleration sensor is also installed on the platform body to monitor the instantaneous acceleration a(t) of the platform. The acceleration a(t) is calculated by detecting the velocity change rate of the platform:

[0024]

[0025] Acceleration data is used by the system to identify whether the platform has abnormal acceleration or deceleration.

[0026] Further, the step S3 includes real-time weight optimization calculation, counterweight adjustment and drive control, and dynamic monitoring and weight adjustment;

[0027] The formula for real-time weight optimization calculation is:

[0028]

[0029] Among them, P optimal Indicates the optimal position of the counterweight, F i (t) is the real-time load or motion data of the i-th sensor at time t, w i(t) is the dynamic weight coefficient of the i-th sensor, reflecting its current impact on the platform balance, and n is the number of sensors;

[0030] The calculation formula for counterweight adjustment is:

[0031]

[0032] Among them, ΔL is the change in chain length that needs to be adjusted, F imbalanced is the unbalanced force caused by uneven load distribution, H is the height difference between the platform and the counterweight, and W counterweight is the total weight of the counterweight,

[0033] The system is based on the unbalanced force F imbalanced Adjust the length ΔL of the chain to ensure that the platform remains stable during the lifting process;

[0034] The dynamic monitoring and weight adjustment formula is:

[0035]

[0036] Among them, w i (t+1) is the weight of the i-th sensor at the next time point, w i (t) is the weight at the current time point, F i (t) is the real-time load value of the i-th sensor, F avg (t) is the current average load value of all sensors, and α is the adjustment coefficient, which controls the dynamic adjustment range of the weight.

[0037] Further, the step S4 includes starting the fuzzy PID control algorithm, adjusting the PID parameters when the load changes, and judging the fuzzy rules when the load fluctuates greatly;

[0038] Starting the fuzzy PID control algorithm includes starting the fuzzy PID controller, which takes the current speed deviation e(t) and acceleration change of the lifting platform as input. The speed deviation e(t) is the difference between the expected speed and the actual speed, which is defined as follows:

[0039] e(t)=v desired (t)-v(t)

[0040] Among them, e(t) is the speed deviation at the current time t, v desired (t) is the expected speed of the platform calculated based on the load and wind speed, and v(t) is the actual lifting speed of the platform, which is provided by the motion sensor;

[0041] Based on the size and change rate of the deviation, the fuzzy controller outputs the adjustment value Δv to correct the current lifting speed. The formula is as follows:

[0042] v(t+1)=v(t)+Δv

[0043] Among them, v(t+1) is the lifting speed of the platform at the next moment, and Δv is the speed adjustment value output by the fuzzy controller according to the deviation;

[0044] The PID parameter adjustment when the load changes includes adjusting the PID controller through the following formula:

[0045]

[0046] Where v(t) is the current lifting speed of the platform, e(t) is the speed deviation, that is, the difference between the expected speed and the actual speed, K p is the proportional coefficient, which is used to linearly control the relationship between speed deviation and output speed, K i K is the integral coefficient, which corrects the speed deviation accumulated over time. d The system dynamically adjusts these PID parameters according to the degree of load change. When the load fluctuates greatly, the fuzzy rule judgment includes judging the optimal lifting speed through the fuzzy rule base of the fuzzy PID algorithm and automatically adjusting the PID gain coefficient.

[0047] Furthermore, step S5 includes starting the deep learning model to continuously run based on the data collected in step S2, and predicting the load distribution of the platform in real time by analyzing the current and historical load changes and environmental data, so as to provide an adjustment basis for the operation of the platform in advance;

[0048] By comparing these real-time data with the model's prediction results, the neural network can identify prediction errors and adaptively adjust the model's parameters based on error feedback to optimize the prediction ability.

[0049] This process is represented as:

[0050]

[0051] Among them, ΔW represents the update of the neural network weights, η represents the learning rate, which controls the step size of the adjustment, and L represents the loss function, which measures the error between the predicted results and the actual results. Represents the partial derivative of the loss function with respect to the weight, indicating the degree of influence of the error on the network weight.

[0052] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: The present invention combines multi-layer neural network and fuzzy PID control technology, and improves the stability and safety of the platform under complex working conditions by dynamically adjusting the lifting speed and counterweight distribution of the platform, and solves the problems of slow response and insufficient adjustment accuracy of traditional methods when dealing with load fluctuations and environmental changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a flow chart of the method of the present invention;

[0054] Figure 2 This is a schematic diagram of the structure of the wind turbine tower lifting platform;

[0055] Figure 3 It is a schematic diagram of the structure of a wind turbine tower lifting platform with an anti-fall component;

[0056] Figure 4 This is an enlarged view of the A-end structure;

[0057] Figure 5 This is an enlarged view of the B-end structure;

[0058] Figure 6 It is a top view of the platform body;

[0059] Figure 7 It is the side view of the platform body;

[0060] Figure 8 It is a multi-layer neural network diagram. DETAILED DESCRIPTION

[0061] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0062] like Figure 2 As shown, the main installation structure of the wind turbine tower lifting platform based on adaptive weight and fuzzy PID control of the present invention. The main frame 1 is the supporting structure of the lifting platform, which is installed on the outside of the offshore wind turbine tower to provide a stable framework for the lifting platform. The ladder component 2 is installed on the main frame to provide a backup manual ladder path to enhance the safety of the platform. The guide rail assembly 3 is arranged along the main frame to ensure the stability of the lifting platform when sliding up and down. The lifting platform mechanism 4 is the core of the entire device and is used for lifting people or cargo. Figure 4 , Figure 5As shown, the connector 41 connects the lifting platform mechanism and the chain to ensure that the lifting platform can cooperate with the drive system. The platform body 42 is used to actually carry the operation and maintenance personnel or items. The platform frame 43 supports the structure of the platform body and connects other related components. The slider 44 cooperates with the guide rail assembly to ensure that the lifting platform slides smoothly on the main frame. The pull-down protection rod 45 provides additional safety protection to prevent the operation and maintenance personnel from accidentally falling. The guard plate 46 protects the platform personnel and prevents the falling objects from affecting the safety of the operation and maintenance personnel. The rotating anti-fall sleeve 47 improves safety and ensures that the platform will not fall when it is accidentally tilted. The hook 48 is used to connect protective equipment such as safety belts to further ensure the safety of the operation and maintenance personnel. The flip stack 49 can be unfolded to place tools for easy use during maintenance operations. The lifting drive assembly 5 is a key component that drives the lifting and lowering movement of the platform. The transmission sprocket assembly 6 is connected to the drive assembly through a chain to drive the lifting platform up and down. The driven sprocket assembly 7 cooperates with the transmission sprocket to ensure that the platform and the counterweight assembly work smoothly. The counterweight assembly 8 keeps the platform balanced and stable during the lifting process. The anti-fall assembly 10 provides emergency protection to prevent the platform from falling when an accident occurs. Figure 6 As shown, the top corridor 11 provides a passage for operation and maintenance personnel to enter the nacelle of the offshore wind turbine tower through the platform. The sliding terminal 13 cooperates with the conductive slide to ensure the stability of the power supply when the lifting platform moves. The conductive slide 14 provides stable power transmission for the platform to ensure the normal operation of the control system. The protective strip 15 increases the safety of the conductive slide to avoid danger during power transmission.

[0063] like Figure 1 The specific adjustment method is as follows:

[0064] S1. Sensor installation and data collection, including platform load, wind speed, platform motion status and other data, to ensure that the collected data can fully reflect the operating status of the platform;

[0065] S2, data processing and analysis;

[0066] S3, adaptive weight optimization adjustment, judges the current platform load and environmental conditions based on the data of S2, and calculates the optimal counterweight position in real time;

[0067] S4, application of fuzzy PID control algorithm, to ensure that the lifting speed matches the current load according to the weight optimization result of S3;

[0068] S5, multi-layer neural network load prediction, predicts load changes in advance by analyzing the real-time lifting status and load information in S4;

[0069] S6. Security monitoring and feedback mechanism to ensure that abnormal situations in all links can be captured in a timely manner.

[0070] In this implementation, S1 includes the following sub-steps:

[0071] S11, collecting load data of the platform;

[0072] First, strain gauge load sensors are installed at four key support points on the bottom of the platform (the four corners of 42). These sensors monitor the load on the platform in real time and convert mechanical force into electrical signals through precise strain detection to capture the load applied to the platform.

[0073] Then, the electrical signal collected by the sensor is transmitted to the central control system through the amplifier. The signal is converted into a digital signal by the analog-to-digital converter (ADC) and input into the control system for data processing.

[0074] The real-time total load Q of the platform is determined by the output signal F of each sensor. i The calculation formula is:

[0075]

[0076] Where Q is the total load of the platform, F i is the load value detected by the ith sensor, and n is the number of sensors installed on the platform (usually 4, distributed at the four corners of the platform).

[0077] Next, in order to improve the accuracy of the load data, the system will weight the sensor signal through a dynamic weight optimization algorithm. The specific formula is:

[0078]

[0079] Among them, Q adjusted is the load data after weight adjustment, w i is the dynamic weight coefficient of the sensor, reflecting the credibility of its data, F i is the load value collected by sensor i, and n is the number of sensors.

[0080] These weight coefficients are adjusted according to real-time operating conditions and historical data to ensure dynamic balance of load data.

[0081] Finally, all sensor data are transmitted to the central control system through the data bus and fed back to the adaptive adjustment module in real time, providing basic data support for subsequent lifting and lowering control.

[0082] S12, collecting wind speed data of the platform;

[0083] First, environmental sensors are installed on the upwind side of the platform (usually the upper part of the platform frame) to monitor wind speed. These wind speed sensors use ultrasonic or mechanical anemometer technology to monitor wind speed changes in the external environment in real time.

[0084] The wind speed sensor converts the detected wind speed value into an electrical signal and transmits it to the central control system for real-time processing after analog-to-digital conversion. wind It is the instantaneous wind speed, usually expressed in metres per second (m / s).

[0085] Next, the system will associate the wind speed data with the platform's operating status to prevent the platform from continuing to operate when the wind speed exceeds the safety threshold. threshold This can be calculated using the following formula, taking into account the height of the platform and the current load:

[0086]

[0087] Among them, V threshold is the wind speed safety threshold (m / s), k is the wind speed safety factor, which is usually determined according to the platform design parameters, H is the height of the platform (m), and Q is the real-time total load on the platform (kg).

[0088] When the wind speed V wind Exceeding the safety threshold V threshold When the wind speed reaches 400°, the system will automatically trigger the safety mechanism and respond in time through the linked anti-fall sleeve 47 and guard plate 46 to prevent the platform from operating under high wind speed.

[0089] Finally, the data from the wind speed sensor will be continuously monitored and combined with the platform's operating status to ensure that the platform operates within a safe wind speed range. Once the wind speed exceeds the threshold, the control system will pause the platform's lifting action and issue a warning to the operator, and if necessary, activate protective devices to ensure safety.

[0090] S13, collecting motion state data of the platform;

[0091] First, motion sensors are installed on the platform frame 43 to detect the lifting speed, acceleration and position change of the platform. Key sensors include position sensors, velocity sensors and acceleration sensors to ensure that the platform can accurately capture its motion state during operation. Position sensors (such as laser rangefinders or encoders) are installed near the platform rails (3) to detect the vertical position x(t) of the platform in real time and calculate the displacement by integrating the velocity v(t):

[0092] x(t)=x0+∫0 t v(t)dt

[0093] Among them, x0 is the initial position of the platform, and v(t) is the instantaneous speed of the platform, which is detected by the speed sensor installed on the drive system 5 or the guide rail assembly 3. These sensors continuously collect the movement speed of the platform relative to the guide rail to ensure that the platform runs smoothly at the set safety speed.

[0094] At the same time, an acceleration sensor is also installed on the platform body 42 to monitor the instantaneous acceleration a(t) of the platform. The acceleration a(t) is calculated by detecting the velocity change rate of the platform:

[0095]

[0096] Acceleration data can help the system identify whether the platform has abnormal acceleration or deceleration. Especially in emergency situations, the system can predict potential dangers through changes in acceleration and take necessary safety measures.

[0097] All the collected motion state data is transmitted to the central control system through the data interface. The system will analyze the data in real time and combine S11 (load data) and S12 (wind speed data) to ensure the stability of the platform operation. If the system detects an abnormal motion state, such as a sudden increase in speed, acceleration exceeding the set threshold, or position deviation exceeding the safe range, the control system will automatically adjust the platform's operating speed or even trigger a pause operation to prevent accidents.

[0098] S2 includes the following sub-steps:

[0099] S21. Data processing and preprocessing

[0100] First, the multi-dimensional data from the sensors, including the platform's load, wind speed, motion status and other information, are transmitted to the central control system through the data interface. The data received by the system may contain noise or outliers, so the data needs to be preprocessed. The preprocessing steps include data cleaning and noise reduction to ensure that the data from all sensors are highly accurate and consistent. The platform's load information is collected through S11, environmental data (such as wind speed) is collected through S12, and motion status data is collected through S13. In order to improve data quality, the system uses a noise reduction algorithm based on Kalman filtering or low-pass filtering:

[0101] D filtered (t) = α·D(t) + (1-α)·D filtered (t-1)

[0102] Among them, D filtered (t) represents the filtered clean data at time t, D(t) is the original sensor data, and α is the filter coefficient, which is used to control the degree of data smoothing.

[0103] After the data is cleaned and filtered, the system can eliminate noise and anomalies, ensuring that subsequent adaptive weight optimization and fuzzy PID control can make decisions based on high-quality data.

[0104] S22. Data storage and analysis

[0105] After data preprocessing is completed, all processed data will be stored in the multidimensional data framework of the central control system. This data framework contains the platform's real-time load data, wind speed data, motion state data, and historical balance data. During the storage process, the system will structure the data according to timestamps and classify them for subsequent algorithm calls and use.

[0106] The system not only stores the current real-time data, but also regularly analyzes the historical operation data of the platform, especially the lifting speed, position and other data obtained from the lifting drive component (5). These historical data are used to support future regulation needs and predict possible load change patterns or wind speed change trends through machine learning or data mining technology, thereby providing a reference for the adaptive weight optimization algorithm. The prediction model can be expressed by the following formula:

[0107] P(t+1)=f(D current , D history )

[0108] Among them, P(t+1) is the predicted value of the next time step, D current is the current real-time data, D history is the historical data, and f is the machine learning or statistical model used for prediction.

[0109] In this way, the system can provide decision support for load regulation and platform operation in advance under complex working conditions to ensure stable operation of the platform.

[0110] S3 includes the following sub-steps:

[0111] S31, real-time weight optimization calculation

[0112] First, based on the load data, wind speed data, and platform motion state data processed and analyzed in S2, the central control system starts the adaptive weight optimization algorithm to calculate the optimal position of the counterweight in real time. At this time, the system will take into account the load changes on the platform, external wind speed fluctuations, and the platform's motion state to ensure that the platform's center of gravity is adjusted through the counterweight to keep the platform running smoothly during the lifting process.

[0113] The core of weight optimization is to dynamically calculate the weight w of each sensor i (t) to reflect the contribution of different sensors to the platform balance. The system performs real-time weight optimization calculations using the following formula:

[0114]

[0115] Among them, P optimal Indicates the optimal position of the counterweight, F i(t) is the real-time load or motion data of the i-th sensor at time ttt, w i (t) is the dynamic weight coefficient of the i-th sensor, reflecting its current impact on the platform balance, and n is the number of sensors.

[0116] These weight coefficients w i (t) is dynamically adjusted and optimized according to the current platform load, wind speed and motion state. The weight adjustment process relies on the platform's real-time data and historical data (such as previous load patterns) to ensure that the counterweight can be optimally adjusted under real-time changing working conditions to prevent the platform from unbalancing.

[0117] As load and environmental conditions continue to change, the system dynamically adjusts the weight w at each moment t i (t) to ensure the balance of the platform. For example, when the load in a certain area increases, the sensor weight in that area will increase, thereby allocating more weight to the corresponding position to keep the platform stable.

[0118] Through this real-time adaptive weight optimization, the system can effectively cope with load changes under complex working conditions, and through dynamic adjustment of the counterweight blocks, ensure that the platform can operate smoothly under various operating conditions.

[0119] S32, counterweight adjustment and drive control

[0120] After the weight optimization calculation in S31 is completed, the system determines the optimal position of the counterweight. Next, the drive system will adjust the balance state of the platform by adjusting the relative position of the chain and the counterweight according to the optimization result. The position adjustment of the counterweight is achieved through the coordinated work of the lifting drive assembly 5 and the transmission sprocket assembly 6.

[0121] First, the central control system sends a control signal to the lifting drive assembly 5, and the drive system starts the motor to control the up and down movement of the counterweight 8 through chain transmission. The transmission sprocket assembly 6 is connected to the lifting platform so that one end of the chain is connected to the counterweight and the other end is connected to the lifting platform. As the system adjusts the counterweight, the length of both ends of the chain changes to maintain the dynamic balance of the platform and the counterweight.

[0122] The specific weight adjustment process is as follows:

[0123]

[0124] Among them, ΔL is the change in chain length that needs to be adjusted, F imbalanced is the unbalanced force caused by uneven load distribution, H is the height difference between the platform and the counterweight, and W counterweight is the total weight of the counterweight.

[0125] The system is based on the unbalanced force Fimbalanced Adjust the length ΔL of the chain to ensure that the platform remains stable during the lifting process.

[0126] During the platform lifting process, the real-time adjustment of the counterweight ensures that the platform will not tilt or become unstable due to load changes or wind speed fluctuations. The movement of the chain and the counterweight is synchronized with the lifting and lowering of the platform. Any movement of the platform is promptly fed back to the counterweight system through the transmission sprocket assembly 6 and the chain for dynamic adjustment.

[0127] Finally, the system continuously monitors the status of the platform to ensure that it is always balanced during the lifting process. The counterweight adjustment is carried out by stacking, that is, the system can increase or decrease the position and number of counterweight blocks according to actual needs, thereby further improving the stability of the platform.

[0128] S33. Dynamic monitoring and weight adjustment

[0129] During the operation of the platform, the system monitors real-time data through a series of sensors (load, wind speed and motion state sensors described in S1) to continuously track the load changes and wind speed fluctuations of the platform. Based on the collected real-time data, the system automatically adjusts the weight dynamically to cope with the impact of environmental changes on the stability of the platform.

[0130] First, the central control system continuously monitors the load data F on the platform i (t) and environmental data (such as wind speed V wind ). When the load distribution on the platform is uneven or the wind speed exceeds the safety threshold V threshold When the system dynamically adjusts the weight w of each sensor according to the adaptive weight optimization algorithm i (t) and recalculate the adjustment plan of the counterweight to ensure that the platform remains stable under real-time changing working conditions.

[0131] The dynamic weight adjustment formula is:

[0132]

[0133] Among them, w i (t+1) is the weight of the i-th sensor at the next time point, w i (t) is the weight at the current time point, F i (t) is the real-time load value of the i-th sensor, F avg (t) is the current average load value of all sensors, and α is the adjustment coefficient, which controls the dynamic adjustment range of the weight.

[0134] When the load or wind speed exceeds the set safety range, the system adjusts the weight distribution of each sensor in real time according to the formula, and calculates the optimal position of the counterweight block according to the latest weight value. The system adjusts the position of the counterweight block through the lifting drive component (5) and the transmission sprocket component (6) to ensure that the platform always remains stable under load changes and wind speed fluctuations.

[0135] The adjustment of the counterweight is carried out by stacking. The system can optimize the balance of the platform by adding or reducing the counterweight according to different load and wind speed conditions. The stacking counterweight method enables the system to flexibly respond to changes under different working conditions and ensure the smooth operation of the platform.

[0136] The process of real-time monitoring and dynamic adjustment is continuous. The system will make instant adjustments based on each monitored load and environmental change to prevent platform instability caused by load offset or wind speed fluctuation. Through this dynamic monitoring and adjustment mechanism, the platform can maintain efficient stability in complex operating environments.

[0137] S4 includes the following sub-steps:

[0138] S41, start fuzzy PID control algorithm

[0139] After S3 completes the optimization adjustment of the counterweight, the system will further control the lifting speed of the platform through the fuzzy PID control algorithm to ensure that it matches the current load and environmental conditions. The fuzzy PID control algorithm makes real-time decisions based on the data processed from the S1 sensor and S2. The system adjusts the speed based on the data of load, wind speed and motion status.

[0140] First, the system starts the fuzzy PID controller based on the counterweight adjustment result in S3 and the real-time load data. The fuzzy PID controller takes the current speed deviation e(t) and acceleration change of the lifting platform as input. The speed deviation e(t) is the difference between the expected speed and the actual speed, which is defined as follows:

[0141] e(t)=v desired (t)-v(t)

[0142] Among them, e(t) is the speed deviation at the current time t, v desired (t) is the expected speed of the platform calculated according to the load and wind speed, and v(t) is the actual lifting speed of the platform, which is provided by the motion sensor.

[0143] The fuzzy controller uses the fuzzy rule base to infer by inputting this deviation and acceleration change. Based on the size and rate of change of the deviation, the fuzzy controller outputs the adjustment value Δv to correct the current lifting speed. The formula is as follows:

[0144] v(t+1)=v(t)+Δv

[0145] Among them, v(t+1) is the lifting speed of the platform at the next moment, and Δv is the speed adjustment value output by the fuzzy controller according to the deviation.

[0146] The advantage of fuzzy control is that it does not rely on precise mathematical models, but is based on fuzzy logic reasoning based on empirical rule bases and current working conditions. For example, in the case of large load changes or sudden changes in wind speed, the fuzzy controller can respond quickly according to the preset fuzzy rules and adjust the lifting speed of the platform to ensure smooth operation of the platform.

[0147] In addition, the fuzzy PID controller combines the historical data processed in S2 to better adapt to complex operating environments. For example, under certain historical operating conditions with uneven load distribution, the system can predict the speed adjustment requirements in advance, thereby adjusting the platform speed in advance during actual operation.

[0148] Through this fuzzy PID control algorithm, the system can ensure that the lifting speed of the platform matches the actual situation when the load, wind speed and motion state change, thereby maintaining the smooth lifting of the platform.

[0149] S42, PID parameter adjustment when load changes

[0150] In S41, the system has started the fuzzy PID control algorithm and adjusted the lifting speed of the platform through the basic control logic. However, when the system detects a significant change in the load of the platform through the sensor in S1, the dynamic adjustment mechanism in S42 will further optimize the parameters of the PID controller to ensure that the lifting speed of the platform can be adaptively adjusted according to the current load status. This dynamic adjustment is a supplement and enhancement to the basic control of S41, ensuring the smooth operation of the platform under complex working conditions.

[0151] First, the system monitors the sensor data in S1 in real time, including the real-time load information provided by the load sensor and the speed change of the motion sensor. When the load of the platform changes significantly, for example, due to fluctuations in the quantity or weight of the transported materials, the fuzzy PID control algorithm determines the operating status of the platform based on the current speed deviation e(t) and dynamically adjusts the parameters of the PID controller (proportional K p , integral K i , differential K d ). The PID controller is adjusted by the following formula:

[0152]

[0153] Where v(t) is the current lifting speed of the platform, e(t) is the speed deviation, that is, the difference between the expected speed and the actual speed, Kp is the proportional coefficient, which is used to linearly control the relationship between speed deviation and output speed, K i K is the integral coefficient, which corrects the speed deviation accumulated over time. d It is the differential coefficient, which is used to predict the changing trend of speed deviation and prevent sudden speed fluctuations.

[0154] The system then dynamically adjusts these PID parameters based on the extent of the load change. When the load suddenly increases, the system will increase the proportional gain K p , so that the platform can more quickly compensate for the speed drop caused by the increase in load. If a load reduction or large fluctuation is detected, the system will adjust the differential gain K d To smooth the transition and prevent speed fluctuations caused by excessive deceleration. At the same time, the integral gain K i It is used to smooth out small, long-term deviations and prevent slow instability of the platform under continuous load changes.

[0155] This adjustment process is closely linked to the data analysis and weight optimization in S2. The historical data processed in S2 provides background information for the system, allowing the fuzzy PID controller to respond to different load changes in a personalized way. The system will use a smaller K when the load changes steadily. d , reducing the impact of frequent adjustments; when the load changes sharply, the system increases K d value to ensure that the lifting speed can quickly adapt to the new load state and avoid instability caused by adjustment lag.

[0156] Through this process of automatically adjusting the PID parameters, the system can respond quickly when the load changes, ensuring that the platform lifting speed matches the current load, thereby avoiding instability problems in system operation and maintaining smooth operation of the platform.

[0157] S43, Fuzzy rule judgment when load fluctuation is large

[0158] In S41, the system starts the fuzzy PID control algorithm to keep the platform moving smoothly under normal load conditions, and S42 further adjusts the PID parameters to cope with real-time changes in load. When the load fluctuates greatly, S43 will use the fuzzy rule base of the fuzzy PID algorithm to determine the optimal lifting speed and automatically adjust the PID gain coefficient to ensure the smooth operation of the platform under complex working conditions.

[0159] First, the system continues to monitor the load and motion data provided by the sensor in S1 in real time based on the PID parameter adjustment in S42. If the load fluctuates violently, the system judges the current load change based on the fuzzy rule base. These fuzzy rules generate an optimal adjustment plan for the lifting speed by fuzzifying the load deviation e(t) and the speed change rate Δv. In this step, the system not only analyzes the current state, but also uses the historical data in S2 to predict possible fluctuation trends.

[0160] Then, the fuzzy PID controller adjusts the proportional gain K according to the judgment result of the fuzzy rule base. p , integral gain K i , differential gain K d Parameters. The adjustment of these gain coefficients ensures that the platform can still respond quickly and adapt itself when facing large load fluctuations. Specifically, the fuzzy control logic outputs the following gain adjustment scheme according to the degree of load fluctuation:

[0161] When the load fluctuates greatly, the system increases K d To speed up response and ensure that the platform can compensate for deviations caused by fluctuations more quickly.

[0162] The system also adjusts K d , used to smooth speed changes and prevent the system from overreacting or vibrating violently when the load fluctuates violently.

[0163] K i It is used to correct long-term deviations and prevent the platform from accumulating small deviations in continuous fluctuations, which may lead to system instability.

[0164] With the support of fuzzy control logic, the adjustment of PID parameters is more flexible, and the system can automatically determine the optimal adjustment value of the lifting speed according to the current load status. The fuzzy PID controller uses the reasoning of fuzzy rules to convert the input complex environmental data into precise adjustment operations, ensuring that the platform can still maintain stable operation under fluctuating conditions.

[0165] In addition, the fuzzy rule judgment in S43 not only relies on real-time data, but also keeps close linkage with the historical data in S2. By analyzing the pattern of historical load fluctuations, the system can foresee possible fluctuations in the future and respond to these changes by adjusting the PID parameters in advance. This makes the system not only passively respond to current load fluctuations, but also able to make a certain degree of prediction and advance adjustment.

[0166] Finally, S43 and S42 form a continuous control chain. In S42, the dynamic adjustment of PD parameters responds to real-time load changes, while S43 further uses fuzzy control rules to determine the optimal lifting speed and optimize gain parameters when the load fluctuates greatly. This process ensures that the platform can operate smoothly and safely under different load fluctuation conditions.

[0167] S5 includes the following sub-steps:

[0168] S51. Based on the data collected in S2, the deep learning model starts to run continuously. By analyzing the current and historical load changes and environmental data, it predicts the load distribution of the platform in real time and provides adjustment basis for the operation of the platform in advance.

[0169] First, the system collects multi-dimensional data from S2, including the real-time lifting status of the platform, load information, environmental data such as wind speed, and historical balance data, as input to the multi-layer neural network model. This model is based on deep learning technology, such as Figure 8 As shown, the system starts to run continuously and analyze the current working conditions and historical load changes in real time.

[0170] In this process, the neural network extracts features and recognizes patterns from input data through a multi-layer structure, extracting key load change patterns. Each layer of the network undertakes different feature extraction tasks. The first layer can capture basic load trends, such as the impact of wind speed on load; the middle layer extracts more complex association patterns, such as the nonlinear relationship between load, platform speed and external environmental changes; the final output layer integrates these complex features into load prediction results, providing predictions of future load changes.

[0171] The neural network model uses the relationship between current data and historical data to make predictions by analyzing the time series change pattern of the load. The prediction formula is as follows:

[0172]

[0173] in, is the predicted load of the platform in the future, L(t), L(t-1) are the actual loads at the current and past moments, E(t), E(t-1) are the current and historical wind speed parameters, W(t), Q(t-1) are the current and historical wave load parameters, and f is the prediction function of the neural network model.

[0174] Through such continuous analysis, the neural network can capture the load distribution trend of the platform in advance. Then, based on these prediction results, the model provides the control system for subsequent adjustment decisions. For example, when the system predicts that the platform load will change dramatically in a short period of time, the output of the neural network will be passed as input to the counterweight adjustment and fuzzy PID control algorithm modules to ensure that the platform can be adjusted in advance.

[0175] S52. When the system predicts through the multi-layer neural network model that the future load is about to change significantly, the system will make adjustments in advance to ensure the stability of the platform. Specifically, when the predicted value of the load shows that the platform will face a large load change in a short period of time (such as a sharp increase or decrease in load), the system will immediately enter the counterweight adjustment process to maintain the balance of the platform.

[0176] First, the system uses the weight optimization algorithm in S2 to calculate the optimal counterweight adjustment plan. According to the load change trend caused by the wind speed and wave load predicted in S51, the system determines the optimal position of the counterweight 8 in advance to ensure that the center of gravity of the platform has adapted to the new load requirements before the load change occurs. The adjustment mechanism of the counterweight is realized through the transmission sprocket assembly 6 and the lifting drive assembly 5. The system adjusts the position relationship between the chain and the counterweight in real time according to the predicted data.

[0177] Secondly, the counterweight adjustment process is closely linked to the weight optimization in S2. The counterweight adjustment is based on the results calculated by the adaptive weight optimization algorithm and dynamically corrected in combination with historical data. The system ensures the accuracy of the counterweight adjustment by real-time monitoring of external environmental parameters such as load, wind speed and wave data. For dynamic load changes under complex working conditions, the system can complete the counterweight adjustment in advance to avoid platform tilt or imbalance due to delayed reaction.

[0178] This adjustment process ensures that the platform can always be in a balanced state when future load changes occur, reducing the risk of system instability caused by load fluctuations. The system uses the predictive ability of multi-layer neural networks and, through linkage with S2 weight optimization, enables the platform to respond to load changes in advance, further improving the stability and security of the platform.

[0179] S53. During the real-time operation of the system, the neural network model does not rely solely on static initial training data, but continuously optimizes its prediction accuracy through continuous feedback learning. This continuous feedback mechanism enables the neural network model to continuously adjust itself according to the data in the actual working conditions, and improve its adaptability to the dynamic load changes of the platform.

[0180] First, after the system adjusts the load according to the prediction results (such as adjusting the counterweight in S52), the actual operation data of the platform will be collected again and fed back to the neural network model. The feedback data not only includes the real-time load, wind speed and other environmental information of the platform, but also includes the adjustment effect made by the system at the previous moment. By comparing these real-time data with the prediction results of the model, the neural network can identify the prediction error and adaptively adjust the parameters of the model according to the error feedback to optimize its prediction ability.

[0181] This process can be expressed by the following formula:

[0182]

[0183] Among them, ΔW represents the update of the neural network weights, η is the learning rate, which controls the step size of the adjustment, and L is the loss function, which measures the error between the predicted results and the actual results. It is the partial derivative of the loss function with respect to the weight, indicating the degree of influence of the error on the network weight.

[0184] Through this adaptive update, the neural network can gradually reduce the prediction error and improve the prediction accuracy of the model. Each feedback learning process will provide a new optimization direction for the neural network model, enabling it to more accurately predict future load changes, especially under complex and changeable working conditions. The system's ability to respond to various load fluctuations will be significantly improved.

[0185] Secondly, the feedback learning of the neural network model also involves the enhancement of adaptability to dynamic environments. As the platform operation time increases, the system may face new external conditions (such as changes in wind speed or changes in platform load patterns). The neural network model gradually improves its adaptability by continuously learning data from these new environments and constantly revising its prediction strategies for different environments and load combinations.

[0186] S6 includes the following sub-steps:

[0187] S61, Real-time monitoring and feedback system

[0188] First, the system monitors the operating status of the platform in real time through a variety of sensors installed in S1. These sensors include load sensors, wind speed environment sensors, and motion sensors, which are used to continuously monitor key data such as the platform's load, lifting speed, tilt angle, and wind speed. These data are transmitted to the central control system through a data interface, and the system conducts a comprehensive evaluation of all real-time data to ensure the normal operation of the platform.

[0189] Once the system detects an abnormal situation, such as the platform load exceeds the safety range, the wind speed suddenly increases, or the platform tilts, the system will immediately trigger the preset safety response mechanism. Specifically, the system will analyze the severity of the abnormal situation through the safety algorithm and take corresponding measures according to different abnormal levels, such as slowing down, stopping the platform lifting or activating protective equipment.

[0190] The real-time monitoring system also has a feedback mechanism. All real-time data is not only used for immediate judgment, but also stored in the system database as historical data for subsequent analysis and improvement. These feedback data can help the system better understand the operating characteristics of the platform in different environments and improve the accuracy and response speed of future monitoring.

[0191] Through this continuous monitoring and feedback mechanism, the system can quickly capture potential risks in platform operation and ensure rapid response when an abnormality occurs to ensure the safety of equipment and personnel. This process forms a closed loop with the data collection in S1, enabling the system to always maintain efficient and safe operation in a dynamic environment.

[0192] S62, anti-fall and guard plate linkage

[0193] During the process of real-time monitoring of the platform operation status by the system, when the platform is detected to be tilted, unbalanced or otherwise unstable, the anti-falling component 9 and the guard plate system 46 will immediately start the linkage mechanism, such as Figure 3 As shown, the safety of the operation and maintenance personnel is ensured. The anti-fall component 9 is responsible for preventing the platform from falling when the platform suddenly slides down or tilts, thereby avoiding the occurrence of serious accidents.

[0194] like Figure 7 As shown, the guard plate 46 and the rotating anti-fall sleeve 47 cooperate with each other to form a safety protection barrier. When the platform tilts, the guard plate will provide additional physical protection to prevent people from sliding off the platform due to inertia or accident. The rotating anti-fall sleeve 47 will automatically lock to ensure that the platform can still provide stable support when tilting and prevent people from falling from the edge of the platform. This design provides multiple protections for operation and maintenance personnel and improves the safety of the platform under abnormal working conditions.

Claims

1. A wind turbine tower lifting platform based on adaptive weight and fuzzy PID control, installed on the main body of an offshore wind turbine tower, characterized in that: The invention comprises a main frame (1), a ladder component (2), a guide rail assembly (3), a lifting platform mechanism (4), a lifting drive assembly (5), a transmission sprocket assembly (6), a driven sprocket assembly (7), a counterweight assembly (8), an anti-fall assembly (9), a top gallery (11), a sliding terminal (12) and a conductive slide (13), wherein the main frame (1) is a supporting structure of the lifting platform and is installed on the outside of the main body of the offshore wind power tower, the ladder component (2) is installed on the main frame (1), and the guide rail assembly (3) is installed on the main frame (1). ) is arranged along the main frame (1), the lifting platform mechanism (4) is arranged on the main frame (1) and cooperates with the guide rail assembly (3) and the ladder component (2), and moves up and down along the main frame (1) for lifting people or objects, and the top of the ladder component (2) is provided with a driving sprocket assembly (6) and a driven sprocket assembly (7), the lifting platform mechanism (4) is connected to the counterweight assembly (8), and the lifting is controlled by increasing or decreasing the weight of the counterweight assembly (8), and the ladder component (2) is also provided with an anti-fall assembly (9).

2. The wind turbine tower lifting platform with adaptive weight and fuzzy PID control according to claim 1 is characterized in that: The lifting platform mechanism (4) comprises a connecting head (41), a platform body (42), a platform frame (43) and a slider (44); the lower end of the guide rail assembly (3) is provided with a connecting head (41) for connecting the lifting platform mechanism (4) and a chain; the lower end of the platform body (42) is provided with a platform frame (43); the platform frame (43) and the slider (44) are connected to move on the guide rail assembly (3).

3. The wind turbine tower lifting platform with adaptive weight and fuzzy PID control according to claim 1 is characterized in that: The side of the platform body (42) is provided with a pull-down protection rod (45), a guard plate (46) and a rotating anti-fall sleeve (47), and the other side is provided with a top corridor (10) and a conductive slide (12), a sliding terminal (11) is provided on the conductive slide (12), a protection strip (13) is provided between the conductive slide (12) and the sliding terminal (11), and a hanging buckle (48) and a flip stacking plate (49) are provided on the top of the platform body (42).

4. A wind power tower lifting platform adjustment method based on adaptive weight and fuzzy PID control, characterized in that: The steps include: S1. Sensor installation and data collection, including platform load, wind speed, platform motion status and other data, to ensure that the collected data can fully reflect the operating status of the platform; S2, data processing and analysis; S3, adaptive weight optimization adjustment, judges the current platform load and environmental conditions based on the data of S2, and calculates the optimal counterweight position in real time; S4, application of fuzzy PID control algorithm, to ensure that the lifting speed matches the current load according to the weight optimization result of S3; S5, multi-layer neural network load prediction, predicts load changes in advance by analyzing the real-time lifting status and load information in S4; S6. Security monitoring and feedback mechanism to ensure that abnormal situations in all links can be captured in a timely manner.

5. The wind power tower lifting platform adjustment method based on adaptive weight and fuzzy PID control according to claim 4 is characterized in that: The step S1 includes: collecting the load data, wind speed data and motion state data of the platform, and transmitting all the collected motion state data to the central control system through the data interface. The system analyzes these data in real time and combines the load data and wind speed data to ensure the stability of the platform operation. If the system detects an abnormality in the motion state, the control system will automatically adjust the operating speed of the platform.

6. The wind power tower lifting platform adjustment method based on adaptive weight and fuzzy PID control according to claim 5 is characterized in that: The load data of the platform is collected by installing strain gauge load sensors at four key supporting points at the bottom of the platform. These sensors monitor the load on the platform in real time and convert mechanical force into electrical signals through precise strain detection and transmit them to the central control system through amplifiers. The signals are converted into digital signals through analog-to-digital converters (ADCs) and input into the control system for data processing. The real-time total load Q of the platform is determined by the output signal F of each sensor. i The calculation formula is: Where Q is the total load of the platform, F i is the load value detected by the i-th sensor, and n is the number of sensors installed on the platform.

7. The wind power tower lifting platform adjustment method based on adaptive weight and fuzzy PID control according to claim 5 is characterized in that: The acquisition of the motion state data of the platform includes installing a motion sensor on the platform frame, near the platform rail, detecting the vertical position x(t) of the platform in real time, and calculating the displacement by integrating the velocity v(t): Among them, x0 represents the initial position of the platform, v(t) represents the instantaneous velocity of the platform, which is detected by the velocity sensor installed on the drive system or guide rail assembly. At the same time, an acceleration sensor is also installed on the platform body to monitor the instantaneous acceleration a(t) of the platform. The acceleration a(t) is calculated by detecting the velocity change rate of the platform: Acceleration data is used by the system to identify whether the platform has abnormal acceleration or deceleration.

8. The wind power tower lifting platform adjustment method based on adaptive weight and fuzzy PID control according to claim 1 is characterized in that: The step S3 includes real-time weight optimization calculation, weight adjustment and drive control, and dynamic monitoring and weight adjustment; The formula for real-time weight optimization calculation is: Among them, P optimal Indicates the optimal position of the counterweight, F i (t) is the real-time load or motion data of the i-th sensor at time t, w i (t) is the dynamic weight coefficient of the i-th sensor, reflecting its current impact on the platform balance, and n is the number of sensors; The calculation formula for counterweight adjustment is: Among them, ΔL is the change in chain length that needs to be adjusted, F imbalanced is the unbalanced force caused by uneven load distribution, H is the height difference between the platform and the counterweight, and W counterweight is the total weight of the counterweight, The system is based on the unbalanced force F imbalanced Adjust the length ΔL of the chain to ensure that the platform remains stable during the lifting process; The dynamic monitoring and weight adjustment formula is: Among them, w i (t+1) is the weight of the i-th sensor at the next time point, w i (t) is the weight at the current time point, F i (t) is the real-time load value of the i-th sensor, F avg (t) is the current average load value of all sensors, and α is the adjustment coefficient, which controls the dynamic adjustment range of the weight.

9. The wind turbine tower lifting platform adjustment method based on adaptive weight and fuzzy PID control according to claim 1 is characterized in that: The step S4 includes starting the fuzzy PID control algorithm, adjusting the PID parameters when the load changes, and judging the fuzzy rules when the load fluctuates greatly; Starting the fuzzy PID control algorithm includes starting the fuzzy PID controller, which takes the current speed deviation e(t) and acceleration change of the lifting platform as input. The speed deviation e(t) is the difference between the expected speed and the actual speed, which is defined as follows: e(t)=v desired (t)-v(t) Among them, e(t) is the speed deviation at the current time t, v desired (t) is the expected speed of the platform calculated based on the load and wind speed, and v(t) is the actual lifting speed of the platform, which is provided by the motion sensor; Based on the size and change rate of the deviation, the fuzzy controller outputs the adjustment value Δv to correct the current lifting speed. The formula is as follows: v(t+1)=v(t)+Δv Among them, v(t+1) is the lifting speed of the platform at the next moment, and Δv is the speed adjustment value output by the fuzzy controller according to the deviation; The PID parameter adjustment when the load changes includes adjusting the PID controller through the following formula: Where v(t) is the current lifting speed of the platform, e(t) is the speed deviation, that is, the difference between the expected speed and the actual speed, K p is the proportional coefficient, which is used to linearly control the relationship between speed deviation and output speed, K i K is the integral coefficient, which corrects the speed deviation accumulated over time. d The system dynamically adjusts these PID parameters according to the degree of load change. When the load fluctuates greatly, the fuzzy rule judgment includes judging the optimal lifting speed through the fuzzy rule base of the fuzzy PID algorithm and automatically adjusting the PID gain coefficient.

10. The wind power tower lifting platform adjustment method based on adaptive weight and fuzzy PID control according to claim 1 is characterized in that: The step S5 includes starting to continuously run the deep learning model based on the data collected in step S2, and predicting the load distribution of the platform in real time by analyzing the current and historical load changes and environmental data, so as to provide an adjustment basis for the operation of the platform in advance; By comparing these real-time data with the model's prediction results, the neural network can identify prediction errors and adaptively adjust the model's parameters based on error feedback to optimize the prediction ability. This process is represented as: Among them, ΔW represents the update of the neural network weights, η represents the learning rate, which controls the step size of the adjustment, and L represents the loss function, which measures the error between the predicted results and the actual results. Represents the partial derivative of the loss function with respect to the weight, indicating the degree of influence of the error on the network weight.

Images