A control optimization method and system for gravity energy storage power generation

Abstract

The application discloses a control optimization method and system for gravity energy storage power generation, and relates to the technical field of energy storage and power generation, which comprises the following steps: dividing a drop shaft of a gravity energy storage system into multiple sections along the height direction, collecting output electric power and position signals under a reference drop working condition, calculating estimated energy loss values of the sections based on the difference between theoretical gravity potential energy and actual power generation energy, and determining a target section with the maximum loss value; forming multiple groups of speed distribution curves inside the target section under the condition that the speeds of other sections remain unchanged, completing whole-trip dropping and measuring whole-trip power generation, calculating loss saving amount to determine an optimal speed distribution scheme; after sequentially optimizing multiple high-loss sections, generating a whole-trip speed distribution curve based on smooth interpolation of adjacent section speed distribution curves. The application combines initial theoretical estimation with later measurement, and is high in efficiency and accurate in improving the overall power generation efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy storage and power generation, and more particularly to a control optimization method and system for gravity energy storage power generation. BACKGROUND

[0002] Gravity energy storage is a physical energy storage form that stores potential energy by lifting heavy objects and recovers energy by lowering the heavy objects. It has the advantages of simple structure, long service life and environmental friendliness. A typical gravity energy storage system lifts heavy objects by a winch or a motor, and controls the lowering of the heavy objects to drive a generator to generate power when energy needs to be released. The energy conversion efficiency of such a system is affected by various factors, including shaft resistance, wind resistance, mechanical friction, uneven guide rail, and braking torque control strategy. The traditional control method does not identify different losses in different sections as a whole, cannot identify the energy loss difference between different well sections, and is also difficult to dynamically optimize the operating curve according to the shaft characteristics.

[0003] In actual operation, the friction, air flow and structural eccentricity of different well sections are different, resulting in significant differences in local energy loss. If a unified speed or braking torque strategy is used, it will cause concentrated local section loss and reduce overall power generation efficiency. Although the loss of each section can be estimated by theoretically integrating potential energy and power, this estimation is affected by factors such as power transmission delay and sensor accuracy, making it difficult to accurately reflect the real energy loss. Therefore, a control method and system are needed that can continuously correct and optimize the operating speed of each section while ensuring system safety, thereby improving overall power generation efficiency. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a control optimization method and system for gravity energy storage power generation to solve the problems mentioned in the background.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0006] A control optimization method for gravity energy storage power generation, comprising the following steps:

[0007] The lowering shaft of the gravity energy storage system is divided into multiple sections along the height direction. When the heavy object is operated in the reference lowering condition, the output electric power and position signals of each section are collected respectively. The estimated energy loss value of each section is calculated based on the difference between the theoretical gravitational potential energy and the actual generated energy, and the target section with the maximum estimated loss value is determined accordingly.

[0008] For the target section, multiple groups of different speed distribution curves are formed within the target section while keeping the operating speed of all other sections unchanged.

[0009] Complete one full-process descent under each of the aforementioned speed distribution curves, measure the total power generation of the entire process, and subtract the total power generation of the baseline descent condition as the loss saving amount.

[0010] The speed distribution curve that maximizes the loss savings is selected as the optimization scheme for the target section.

[0011] After removing the optimized target segments, the selection steps of the velocity distribution curve are repeated in the remaining segments in descending order of estimated loss value, until the preset number of high-loss segments are optimized.

[0012] For sections that did not participate in the actual measurement optimization, smooth interpolation is performed based on the velocity distribution curves of adjacent optimized sections to generate the velocity distribution curve of the entire process, and the lowering process of the heavy object is controlled based on the velocity distribution curve of the entire process.

[0013] In some embodiments, the method for estimating energy loss includes the following steps:

[0014] Collect the output power of the heavy object during the lowering process in one section. With displacement The theoretical gravitational potential energy of the calculation section :

[0015] , where m is the mass of the object and g is the acceleration due to gravity;

[0016] At the start time of the corresponding segment and the end time Between output power By integrating, we obtain the estimated actual power generation energy:

[0017] ;in t It is a time variable;

[0018] The difference between the two This serves as the estimated energy loss value for the corresponding section.

[0019] In some embodiments, the displacement of the weight The output power is obtained by measuring the position encoder or laser rangefinder installed inside the wellbore. The output voltage and current are measured in real time by voltage and current sensors of the generator connected to the heavy object and multiplied together.

[0020] In some embodiments, speed regulation is achieved by adjusting the electromagnetic braking torque generated by a generator connected to the weight.

[0021] In some embodiments, when generating the speed distribution curve of the target section, the control system first keeps the running speed at both ends of the section consistent with the speed in the benchmark lowering operation, gradually increases / decreases the lowering speed of the weight in the middle of the section to a target speed corresponding to a set speed increase percentage by reducing / increasing the electromagnetic braking torque of the generator connected to the weight, and then gradually reduces / increases the lowering speed of the weight by increasing / decreasing the electromagnetic braking torque of the generator until the running speed at the other end of the section is again the same as the speed in the benchmark lowering operation, thereby forming a smoothly changing speed distribution curve in the section; and different speed distribution curves are formed by controlling the speed increase / decrease percentage.

[0022] In some embodiments, the benchmark lowering operation is free fall motion under a preset electromagnetic braking torque.

[0023] In some embodiments, the method further comprises recording the total trip power generation and the energy loss change under different speed adjustment amplitudes, and modeling the relationship between the speed adjustment amplitude and the total trip energy loss by using regression analysis or machine learning algorithms to predict the speed increase or decrease percentage that minimizes the total trip energy loss.

[0024] In some embodiments, the method further comprises connecting the speed distribution curves of the sections after obtaining the optimal speed distribution curves of all the sections, and further performing overall fitting processing to generate a continuous and derivable target speed curve over the entire wellbore by using a continuous function approximation method such as polynomial fitting, spline interpolation, or Gaussian process regression.

[0025] In some embodiments, the method further comprises using a speed closed loop The control method acquires the deviation of the lowering speed of the weight from the target speed curve in real time, inputs the deviation into a proportional, integral, and differential link to calculate a torque control amount, converts the control amount into an adjustment instruction for the stator current of the generator, controls the electromagnetic braking torque of the generator by adjusting the size of the stator current, and thereby enables the lowering speed of the weight to stably track the target speed curve.

[0026] The present application also discloses a control optimization system for gravity energy storage power generation, comprising the following modules:

[0027] A section division and data acquisition module is used to divide the lowering wellbore of the gravity energy storage system into multiple sections along the height direction, and acquire the output electric power and position signals of each section when the weight is operated in the benchmark lowering operation.

[0028] An energy loss calculation module is used to calculate and estimate the energy loss value based on the difference between the theoretical gravitational potential energy and the actual power generation energy of each section, and determine the target section with the maximum estimated loss value.

[0029] A torque control module is configured to control the generator connected with the weight to generate an adjustable electromagnetic braking torque under the condition that the running speed of all other sections except the target section is kept unchanged, and to form a plurality of speed distribution curves in the target section by adjusting the size of the electromagnetic braking torque;

[0030] An energy measurement and optimization selection module is configured to complete a whole trip lowering under each of the speed distribution curves, measure the whole trip power generation and compare it with the whole trip power generation of the benchmark lowering working condition, calculate the loss saving amount, and select the speed distribution curve with the highest loss saving amount as the optimization scheme of the target section;

[0031] An iterative optimization module is configured to repeat the optimization selection step in the remaining sections in order from high to low according to the estimated loss value after excluding the optimized target section, until the optimization of a preset number of high loss sections is completed.

[0032] A speed curve generation module is configured to perform smooth interpolation according to the speed distribution curves of adjacent optimized sections, generate a continuous lowering speed distribution curve for the whole wellbore, and output to an execution control unit to realize optimized lowering control.

[0033] The advantages of the present application over the prior art are that the present application proposes a gravity energy storage power generation control optimization method based on segmented estimation combined with whole trip measurement. First, the energy loss values of different sections are estimated by theoretical calculation to quickly lock the key sections with high loss. Although the accuracy of estimation is not high, it can be used for preliminary rapid investigation. Then, under the condition that the operating parameters of other sections remain unchanged, a plurality of speed distribution experiments are performed on the target section, and the loss saving amount brought by the speed adjustment of a single section is evaluated by measuring the change of the whole trip power generation. Since the conditions of other sections are fixed, the loss saving amount of the whole trip power generation is the loss saving amount of the target section, and the actual measurement of the loss saving amount is more accurate than the estimation. The present application combines the initial theoretical estimation with the later measurement verification, which not only ensures the optimization efficiency, but also significantly improves the accuracy of energy loss identification.

[0034] Further, the present application uses the adjustment of the electromagnetic braking torque of the generator to realize speed control, which can smoothly adjust the lowering speed of the weight at the electrical level without additional mechanical components, and has the advantages of fast response and strong controllability. By setting the speeds at both ends to be consistent with the benchmark and adjusting only the speed in the middle section, a speed distribution curve with local speedup or speed reduction can be formed under the premise of maintaining the continuity of the whole trip dynamics, ensuring that the energy optimization process is safe and controllable. In addition, by modeling the relationship between the speed adjustment amplitude and the energy loss change using regression analysis or machine learning algorithm, self-learning and adaptive speed optimization control can be further realized. Finally, the optimal speed distribution curves of all sections are continuously processed by fitting and interpolation technology, and the speed closed loop The control realizes accurate speed tracking and torque coordination control, so that the intelligent lowering and power generation process with high efficiency and low loss is realized under the condition of simple structure.

[0035] The application not only solves the problem of inaccurate energy loss evaluation in the traditional gravity energy storage system, but also proposes a dynamic optimization mechanism that can be iterated step by step and driven by actual measurement feedback, which can maximize the whole-range energy feedback efficiency within a safe range. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 is the overall schematic diagram of the application;

[0037] Figure 2 is the estimation and collection schematic diagram of the application;

[0038] Figure 3 is the target section speed distribution schematic diagram of the application;

[0039] Figure 4 is the whole-range target speed curve and execution control schematic diagram of the application. DETAILED DESCRIPTION

[0040] The specific embodiments of the application will be described below in conjunction with the accompanying drawings.

[0041] The application is directed to the lowering and power generation process of a gravity energy storage device, and proposes a control optimization idea based on sections and actual measurement feedback. The wellbore has friction, wind resistance, guiding deviation and temperature difference in the height direction, which leads to inconsistent energy loss in different sections. The section loss obtained by simply relying on theoretical potential energy and power integration can only be used as a rough estimate, and it is difficult to accurately indicate the real high-loss source due to the influence of sensing accuracy, dynamic inertia and time delay.

[0042] Therefore, as shown in FIG. 1, the application first quickly locates by estimation value, and then gradually approaches the optimal running track by whole-range actual measurement correction, which not only reduces the calculation burden, but also ensures the accuracy of optimization.

[0043] In specific embodiments, first, the lowering wellbore is divided into multiple sections along the height direction, for example, for a 100-meter wellbore, it can be divided into 10-meter sections, a total of 10 sections. During the operation of the reference lowering condition of the weight, the output electric power and position signals of each section are collected. For any section, the theoretical gravitational potential energy of the section is calculated according to the displacement increment and the weight mass and gravitational acceleration, and the estimated actual power generation energy is obtained by time integration of the output power between the start and end time of the section. The difference between the two is the estimated energy loss value of the section.

[0044] Specifically, as shown in FIG. 2, the method for estimating the energy loss value includes the following steps:

[0045] Collecting output power during a drop of a weight in a section and displacement , calculating the theoretical gravitational potential energy of the section representing output;

[0046] where m is the mass of the weight and g is the acceleration of gravity

[0047] Integrating the output power between the start time and the end time of the corresponding section gives the estimated actual energy generated:

[0048] where t is the time variable t

[0049] The difference between the two is taken as the estimated energy loss value of the corresponding section. loss represents loss. By comparing the estimated results of all sections, the target section with the largest estimated loss value can be quickly determined as the priority optimization object. Displacement can be obtained by a position encoder or a laser ranging sensor in the wellbore, and output power can be obtained by multiplying the real-time measurement of the voltage and current on the generator side connected to the weight, so that high-frequency collection and online calculation can be completed without increasing the complexity of the structure. In some embodiments, the reference drop condition can use free fall drop under a preset electromagnetic braking torque.

[0050] As shown in

[0051] , for the determined target section, only form a plurality of different speed distribution curves inside the target section under the condition that the running speed of all other sections outside the target section remains unchanged. The construction of the speed distribution follows the constraint that the speed at both ends is consistent with the reference, and the speed in the middle can be moderately increased or decreased to form a plurality of comparable schemes. In actual execution, the drop speed is adjusted by adjusting the electromagnetic braking torque of the generator connected to the weight, and reducing the braking torque can increase the speed in the middle, and increasing the braking torque can decrease the speed in the middle, and then reverse adjustment is made to make the speed at the end of the section return to consistent with the reference, so as to ensure the continuity of the boundary. In order to form a differentiated scheme, different speed increase percentages or speed decrease percentages can be set to obtain a set of representative speed distribution curves. Figure 3

[0052] ​​​In further embodiments, for each candidate speed profile, a full-length drop is performed and the full-length power generation is directly measured, and the difference between the full-length power generation of the reference drop condition and the full-length power generation is obtained to obtain the loss saving amount. Since the speed of other sections except the target section remains unchanged, the change in full-length power generation can be considered to be caused by the speed adjustment of the target section, thereby avoiding the interference of the estimation error of the section on the judgment. After selecting the speed profile with the highest loss saving amount as the optimization scheme of the target section, it is removed from the unoptimized set, and then a new target section is selected in order of high to low according to the estimated loss value, and the above steps are repeated until a predetermined number of high-loss section optimizations are completed. For those sections that have not been measured and optimized, the speed distribution curve of the adjacent optimized section is interpolated to obtain a continuous and boundary consistent speed distribution curve. The interpolation can be selected from the polynomial interpolation in the prior art.

[0053] In further embodiments, in order to quickly converge to the optimal speed distribution curve with as few test times as possible, improve the parameter adjustment efficiency and the stability of system operation, the system can perform statistical and modeling analysis on the accumulated historical data after multiple drop tests. Each test can record multiple groups of data, such as the speed adjustment amplitude (for example, the speed increase percentage or the speed decrease percentage), the target section coordinate position and length, the start and end speed, the average temperature, the guide rail friction coefficient, the generator rated torque, the electromagnetic braking torque adjustment range, and the final energy loss change amount. These data are regarded as sample points, and the system learns the functional relationship between the speed adjustment and the energy loss through regression modeling or machine learning methods.

[0054] In the initial stage, the system can use polynomial regression or piecewise spline regression to fit the relationship between the speed adjustment amplitude and the full-length energy loss change amount. Polynomial regression can describe the loss change trend caused by speed adjustment, such as the nonlinear characteristics that the loss increases when the speed is too high. Piecewise spline regression can ensure that the fitted curves of different speed intervals are smoothly connected, thereby obtaining a response model that is more consistent with the physical law. For scenes with strong data dispersion and high noise, a Gaussian process regression model can be introduced to define the similarity relationship between the speed adjustment amplitudes through a kernel function, which can not only output the predicted mean value, but also output the predicted variance, so that the system can identify the uncertainty of the prediction. When the model detects that the uncertainty of a certain section is high, it will automatically select a more conservative adjustment amplitude to ensure the safety of operation.

[0055] After the data samples gradually increase, the system can further introduce a machine learning method to realize self-learning. By taking the speed adjustment amplitude, section parameters, and environmental characteristics as input features, and taking the whole-trip energy loss change as an output label, a nonlinear mapping model is trained to capture the energy loss law under the coupling of multiple factors. For example, an integrated learning model such as random forest or gradient boosting tree can be used to evaluate the importance of input features and identify key factors affecting energy loss changes, such as friction, air resistance, or temperature gradient, thereby guiding subsequent optimization strategies. After the model is trained, the system can automatically predict the speed increase or decrease percentage that is most likely to achieve the optimal energy recovery effect in a certain section based on historical data, and use the prediction result as the initial parameter for the next test, thereby reducing the number of manual parameter adjustments.

[0056] In a higher-level embodiment, the system can introduce reinforcement learning or adaptive optimization algorithms, regarding each speed adjustment as a decision action and taking the whole-trip energy loss saving as a reward signal. Through multiple rounds of interactive learning, the control strategy gradually converges to an optimal strategy that can automatically select the appropriate speed adjustment amplitude in different sections. This strategy has the ability to update itself, and when the device state or environmental conditions change, it can real-time correct the model parameters based on a new round of data to realize dynamic learning and continuous optimization.

[0057] In a further embodiment, when the optimal speed distribution of all sections is determined through testing or algorithm prediction, these discrete speed data points can be further integrated to generate a continuous and derivable target speed curve in the entire wellbore height direction. This process can further ensure the smoothness and physical reasonableness of the speed change at the control level, avoiding mechanical impact, electromagnetic fluctuations, or control system oscillation caused by sudden changes in speed between sections, thereby improving the running stability and energy feedback efficiency of the system.

[0058] In a specific implementation, the system first represents the optimal speed distribution of each section as a series of discrete sampling points, each point containing the corresponding wellbore height position and the target speed value at that position. Since different sections may come from different experimental schemes or optimization results, there may be slight discontinuity or derivative changes at the section boundaries. In order to achieve continuity of speed and acceleration throughout the entire section, the system performs continuous function approximation on these discrete data points.

[0059] When the number of wellbore sections is small and the velocity change trend is relatively gentle, a polynomial fitting method can be used to fit an overall polynomial curve using the least squares principle. Polynomial fitting has low computational cost and a simple form, making it suitable for scenarios with a relatively linear overall change trend. However, when the number of sections is large or the curve fluctuates significantly, the "Runge phenomenon" may occur, i.e., oscillations occur at the boundaries. Therefore, in complex multi-section situations, spline interpolation can be used. Spline interpolation is based on piecewise polynomials, forcing the continuity of the first and second derivatives at the connection points of each section, thereby achieving a smooth transition at both the velocity and acceleration levels. This not only eliminates abrupt velocity changes between different sections but also ensures the continuity of force on the mechanical system, helping to reduce guide rail friction impact and generator load fluctuations.

[0060] In a further embodiment, for situations where noise exists or velocity data in some sections is not entirely reliable, Gaussian process regression can be used to generate the target velocity curve. Gaussian process regression is a nonparametric statistical method whose core idea is to treat the velocity distribution as a stochastic process defined by a covariance function, and to obtain a smooth prediction over the entire process by inferring the joint distribution of training samples. This method can not only obtain the predicted velocity value at each location, but also provide the range of uncertainty in the prediction, thereby providing a confidence boundary at the control level. The system can automatically reduce the rate of change of velocity or the amplitude of acceleration in high uncertainty regions to avoid oscillations or overshoot caused by prediction errors.

[0061] Furthermore, such as Figure 4 As shown, the execution control system employs a speed closed-loop approach to achieve precise tracking and real-time adjustment of the load's lowering speed. During operation, the system continuously acquires the actual lowering speed signal of the load and compares it with a real-time reference value of the target speed curve, calculating the deviation between the two. This deviation signal serves as the input to the control input and is fed into the proportional-integral-derivative (PID) circuit. The regulation process performs real-time calculations. The proportional winding generates a correction torque proportional to the current speed deviation to achieve rapid response; the integral winding eliminates steady-state deviations by accumulating historical errors, enabling the system to accurately maintain the target speed during long-term operation; and the derivative winding predicts future trends based on the error change rate, thus applying corrections in advance to suppress overshoot when speed changes or load disturbances occur. The combined effect of these three elements outputs a smooth, dynamic torque control quantity, used to adjust the electromagnetic braking torque generated by the generator in real time.

[0062] The torque control quantity of the control output is then calculated and mapped by the power electronic conversion module to convert into the regulating instruction of the generator stator current. The generator usually adopts a four-quadrant frequency converter or a servo drive system to realize accurate control of the stator current. The control system indirectly changes the combined effect of the armature back electromotive force and the magnetic field interaction by adjusting the amplitude and phase of the stator current, thereby realizing dynamic adjustment of the electromagnetic braking torque. When the amplitude of the stator current increases, the electromagnetic braking torque also increases, so that the speed of the heavy object lowering slows down; when the amplitude of the stator current decreases, the electromagnetic braking torque decreases, and the speed of the heavy object lowering increases. The process is executed in a millisecond cycle, so that the generator can not only maintain stable energy output, but also respond to changes in the target speed curve in real time.

[0063] During the entire closed-loop control process, the controller monitors key variables such as current fluctuation rate, torque change rate, speed error integral value, etc. to evaluate the control quality. If an abnormality is detected, such as a sustained deviation from the target curve or an excessively large oscillation amplitude, the system will automatically adjust parameters, such as reducing the proportional gain to prevent overshoot or increasing the integral time constant to suppress accumulated error. In addition, the system can combine adaptive or fuzzy algorithms to dynamically correct control parameters according to real-time operating conditions and motor temperature changes, so that the control system can maintain optimal response performance under different loads and different environmental conditions.

[0064] Through the above control mechanism, the electromagnetic braking torque of the generator is continuously and flexibly adjusted, and the speed of the heavy object lowering can stably track the pre-generated target speed curve. This closed-loop control method not only avoids the problems of unstable speed and large energy output fluctuations in traditional open-loop control, but also significantly reduces the impact load and structural fatigue of the mechanical system. Ultimately, optimal feedback of power generation and long-term stability of system operation are achieved.

[0065] The present application has the characteristics of explainable, verifiable and reproducible in the optimization process, and realizes precise management of the high-loss section while ensuring the safety boundary, so as to ultimately achieve the goal of improving the whole-range feedback energy, reducing local loss and prolonging the service life of key components. The above is only the preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can make equivalent replacement or change according to the technical solution and inventive concept of the present application within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A control optimization method for gravity energy storage power generation, characterized by, The method comprises the following steps: The drop shaft of the gravity energy storage system is divided into multiple sections in the height direction, and the output electric power and position signals of each section are collected when the weight is operated in a reference drop condition; the estimated energy loss value of each section is calculated based on the difference between the theoretical gravity potential energy and the actual power generation energy, and the target section with the maximum estimated loss value is determined; For the target section, different speed distribution curves are formed inside the target section while keeping the operating speed of all other sections unchanged; The whole drop is completed under each speed distribution curve, the power generation amount of the whole drop is measured, and the loss saving amount is obtained by subtracting the whole drop power generation amount of the reference drop condition; The speed distribution curve with the highest loss saving amount is selected as the optimization scheme of the target section; After the optimized target section is removed, the speed distribution curve selection step is repeated in the remaining sections from high to low according to the estimated loss value until the optimization of the preset number of high-loss sections is completed; For the sections that do not participate in the actual measurement and optimization, the speed distribution curve of the whole drop is generated based on the smooth interpolation of the speed distribution curves of the adjacent optimized sections, and the drop process of the weight is controlled based on the speed distribution curve of the whole drop.

2. The control optimization method of gravity stored energy electric power generation according to claim 1, wherein, The method for calculating the estimated energy loss value comprises the following steps: Collecting the output power of a weight during lowering in a section with displacement , calculating the theoretical gravitational potential energy of the section : where m is the mass of the weight and g is the acceleration due to gravity. The integral of the output power between the start time of the corresponding segment and the end time gives an estimate of the actual generated energy : ; wherein t is a time variable; The difference between the two as an estimated energy loss value for the corresponding section.

3. The control optimization method of gravity stored energy electric power generation according to claim 2, wherein, Displacement of the weight Output power Output voltage and current are measured in real time by voltage and current sensors of the generator connected to the weight and multiplied.

4. The control optimization method of gravity stored energy electric power generation according to claim 1, wherein, The adjustment of the speed is realized by adjusting the electromagnetic braking torque of the generator connected to the weight.

5. The control optimization method of gravity stored energy electric power generation according to claim 4, wherein, When generating the speed distribution curve of the target section, the control system first makes the operating speed at both ends of the section consistent with the speed under the reference drop condition, reduces or increases the electromagnetic braking torque of the generator connected to the weight to gradually increase or decrease the weight drop speed in the middle of the section to the target speed corresponding to the set speed increase percentage, and then increases or reduces the electromagnetic braking torque of the generator to gradually reduce or increase the weight drop speed until the operating speed at the other end of the section is again the same as the speed under the reference drop condition, thereby forming a smooth speed distribution curve in the section; different speed distribution curves are formed by controlling the speed increase or decrease percentage.

6. The control optimization method of gravity stored energy electric power generation according to claim 5, wherein, The method further comprises recording the whole drop power generation amount and energy loss change amount under different speed adjustment amplitudes, and modeling the relationship between the speed adjustment amplitude and the whole drop energy loss by using regression analysis or machine learning algorithm to predict the speed increase or decrease percentage that minimizes the whole drop energy loss.

7. The control optimization method of gravity stored energy electric power generation according to claim 1, wherein, The method further comprises connecting the speed distribution curves of each section after obtaining the optimal speed distribution curves of all sections, and further performing overall fitting processing to generate a continuous and derivable target speed curve on the whole drop shaft by using a continuous function approximation method such as polynomial fitting, spline interpolation or Gaussian process regression.

8. The control optimization method of gravity stored energy electric power generation according to claim 7, wherein, The method further comprises: using a speed closed-loop PID control method, collecting the deviation of the heavy object lowering speed from the target speed curve in real time, inputting the deviation into a proportional, integral, and differential link to calculate a torque control amount, converting the control amount into an adjusting instruction of the generator stator current, controlling the electromagnetic braking torque of the generator by adjusting the size of the stator current, so that the heavy object lowering speed stably tracks the target speed curve.

9. A control optimization system for gravity energy storage power generation, characterized by, The method comprises the following modules: A section division and data collection module is configured to divide the lowering shaft of the gravity energy storage system along the height direction into multiple sections, and collect the output electric power and position signals of each section when the heavy object is operated in the reference lowering condition; An energy loss calculation module is configured to calculate and estimate the energy loss value based on the difference between the theoretical gravity potential energy and the actual power generation energy of each section, and determine the target section with the maximum estimated loss value; A torque control module is configured to control the generator connected to the heavy object to generate an adjustable electromagnetic braking torque while keeping the operating speed of all other sections except the target section unchanged, and form multiple sets of speed distribution curves in the target section by adjusting the size of the electromagnetic braking torque; An energy measurement and optimization selection module is configured to complete one whole course lowering under each speed distribution curve, measure the whole course power generation, compare the whole course power generation with that of the reference lowering condition, calculate the loss saving amount, and select the speed distribution curve with the highest loss saving amount as the optimization scheme of the target section; The reference lowering condition is free fall motion under a preset electromagnetic braking torque; An iterative optimization module is configured to repeat the optimization selection step in the remaining sections in descending order of the estimated loss value after excluding the optimized target section, until the optimization of a preset number of high loss sections is completed; A speed curve generation module is configured to perform smooth interpolation according to the speed distribution curves of adjacent optimized sections, generate a continuous lowering speed distribution curve for the whole shaft, and output the curve to an execution control unit to realize optimized lowering control.

Images