Mining vehicle paying-out and recovering method and system based on umbilical cable slack prediction mechanism

Abstract

The present application relates to the technical field of ocean engineering, and discloses a mining vehicle laying and recovering method and system based on umbilical slack prediction mechanism, which can predict the slack risk in advance and ensure the reliability of intelligent response. The method comprises the following steps: inputting the preprocessed real-time heave data into a heave data prediction model to obtain a heave motion prediction result of the ship stern in the future target time, so as to predict the umbilical slack length D at the next moment in advance; judging whether D is still in a slack state after considering the passive compensation range corresponding to the current sea state; if yes, starting the active compensation system to collect the cable in advance, and the cable length is equal to the length D; otherwise, acquiring real-time umbilical tension data and comparing it with a preset tension threshold value, if the real-time tension exceeds the tension threshold value, starting the active compensation system to execute the winch cable laying adjustment action, and when the umbilical tension data decreases to the tension threshold value range, switching the active compensation system to a dormant state.

Description

Technical Field

[0001] This invention relates to the field of marine engineering technology, and in particular to a method and system for deploying and recovering mining vehicles based on an umbilical cable slack prediction mechanism. Background Technology

[0002] With the gradual advancement of deep-sea mineral resource development, deep-sea mining vehicles, as core equipment for seabed mineral extraction, rely on umbilical cable winch systems for deployment and retrieval. The umbilical cable, with its multiple functions of load-bearing, power transmission, and signal communication, is a crucial link connecting surface mining vessels and underwater mining vehicles. Deep-sea operations can reach depths of 4000–6000 meters. Mining vessels are affected by waves and currents, resulting in vertical heave, pitching, and rolling changes. These movements, transmitted to the umbilical cable, cause severe fluctuations in cable tension, leading not only to underwater instability of the mining vehicle and reduced deployment and retrieval accuracy, but also to umbilical cable slack, twisting, overload breakage, and even damage to the internal electrical / optical transmission units, significantly increasing operational safety risks and equipment wear and tear costs.

[0003] To offset the negative impacts of mining vessel movement, heave compensation technology has become a core support for deep-sea cable operations. Currently, the mainstream technologies are divided into three categories: active, passive, and semi-active. Passive heave compensation systems rely on energy storage components such as hydraulic accumulators and springs to passively buffer tension impacts. They are simple in structure and have low energy consumption, but their compensation response is lagging and their accuracy is low, making them unsuitable for the heavy-load and high-sea-state operation requirements of deep-sea mining vehicles. Active heave compensation systems collect hull motion data through motion reference units (MRUs) and drive actuators to actively offset heave displacements. They offer high compensation accuracy and strong adaptability, but the systems are complex, energy-intensive, and lack the ability to predict umbilical cable slack, making them prone to instantaneous impact loads due to slack followed by re-tensioning. Semi-active heave compensation systems combine the advantages of active and passive systems, using passive buffering as a foundation and active control for correction. This represents the development trend of deep-sea operations, but existing solutions still have many technical defects and cannot directly adapt to the working conditions of deep-sea mining vehicles.

[0004] When existing heave compensation technology is applied to the deployment and retrieval system of umbilical cables for deep-sea mining vehicles, the following main shortcomings exist:

[0005] I. Existing semi-active systems are mostly simple superpositions of active and passive modules, using a fixed threshold mode switching strategy. They do not dynamically allocate the compensation range according to the sea state level, resulting in insufficient compensation accuracy under high sea states and energy waste under low sea states, failing to balance compensation performance and energy economy.

[0006] Second, there is a lack of umbilical cable slack prediction mechanism. The systems designed in existing patent applications such as CN108116623 A and CN119329689A passively respond to slack by real-time data from tension sensors. They cannot predict the risk of slack in advance based on the movement trend of the mining vessel, which makes it impossible to effectively avoid the instantaneous impact load caused by slack, becoming the main cause of umbilical cable damage.

[0007] Third, the level of automation is low. The deployment and retrieval of deep-sea mining vehicles are heavy-duty operations in deep water. It is difficult to manually control the winch in harsh sea conditions. However, the existing system lacks an automated control method adapted to the working conditions of mining vehicles, resulting in low operating efficiency and high dependence on manual labor. Summary of the Invention

[0008] The purpose of this invention is to disclose a method and system for deploying and recovering mining vehicles based on an umbilical cable slack prediction mechanism, so as to predict slack risks in advance and ensure the reliability of intelligent response.

[0009] To achieve the above objectives, the present invention discloses a method for deploying and recovering mining vehicles based on an umbilical cable slack prediction mechanism, comprising: Step S1: Acquire real-time heave data generated by the waves at the stern working end monitored by the attitude detection sensor, input the preprocessed real-time heave data into the heave data prediction model, and obtain the prediction result of the heave motion of the stern in the future target time. Step S2: Based on the heave motion prediction results, combined with the winch cable length, mining vehicle speed and current underwater depth, predict the umbilical cable slack length D at the next moment. Step S3: Determine whether the slack length D of the umbilical cable is still in a slack state after considering the passive compensation range corresponding to the current sea state; if yes, proceed to step S4; otherwise, proceed to step S5. Step S4: Activate the active compensation system and perform the winch cable retrieval action in advance. The retrieved cable length is equal to the slack length D of the umbilical cable. Step S5: Obtain real-time tension data of the umbilical cable and compare it with a preset tension threshold. If the real-time tension exceeds the tension threshold, start the active compensation system to perform winch cable release adjustment. When the umbilical cable tension data decreases to within the tension threshold range, switch the active compensation system to sleep mode.

[0010] Preferably, before step S1, the method further includes: reconstructing training samples of the heave data prediction model using a sliding window based on the historical heave data of the attitude detection sensor in the operating sea area.

[0011] Preferably, the heave data prediction model uses the XGBoost algorithm for initial model construction and incremental learning training; wherein, the attitude detection sensor collects data 10 times per second, and the training samples for reconstructing the heave data prediction model through a sliding window are specifically as follows: the average heave data within one second is taken as the data source for training operations; the heave data for every 5 consecutive seconds is used as feature values, and the heave data for the next 3 seconds is used as labels, so as to construct supervised learning samples.

[0012] Preferably, the sea state is divided into Level 1, Level 2, Level 4, and Level 4 and above; the passive compensation system is in a closed state when the sea state is Level 1, and the corresponding passive compensation range is 0; the passive compensation system is activated when the sea state is at least Level 2.

[0013] Preferably, the passive compensation system comprises: a balance wheel and a hydraulic cylinder fixed to the deck by a hinge; the extension and retraction of the hydraulic cylinder can change the rotation amplitude of the balance wheel; one end of the accumulator is connected to the hydraulic cylinder, and the other end is connected to a low-pressure gas cylinder and a high-pressure gas cylinder respectively via a low-pressure side solenoid valve and a high-pressure side solenoid valve; the method further comprises: when the sea state is identified as level 2 based on a series of continuous real-time heave data, opening the low-pressure side solenoid valve and closing the high-pressure side solenoid valve; when the sea state is identified as at least level 4 based on a series of continuous real-time heave data, closing the low-pressure side solenoid valve and opening the high-pressure side solenoid valve.

[0014] Preferably, the formula for calculating the slack length D of the umbilical cord is: Where L0 is the length of the umbilical cable at the moment the mining vehicle enters the water, L1 is the current length of the umbilical cable, k is the deformation ratio of the umbilical cable with length, L3 is the distance the ship will descend at the next moment as predicted by the heave data prediction model, t is the time interval between adjacent calculation frequencies, H is the underwater depth of the mining vehicle at the current moment, and v is the vehicle's azimuth velocity obtained by the mining vehicle's inertial measurement unit.

[0015] To achieve the above objectives, this invention also discloses a mining car deployment and retrieval system based on an umbilical cable slack prediction mechanism, comprising: an A-frame for stern deployment mounted on the deck of the mining car, a winch with tension and cable length monitoring functions, a pendulum wheel for passive compensation, an attitude sensor for monitoring real-time heave data generated by waves at the stern working end, and an umbilical cable connected thereto; the active compensation winch can actively retrieve or release the cable based on control signals to balance the tension of the umbilical cable, and the passive compensation pendulum wheel can swing within a small range when the tension changes to passively compensate for the tension of the umbilical cable; the mining car is equipped with an inertial measurement unit and a depth sensor; a controller located in the control room establishes signal connections with the attitude sensor, the control unit of the active compensation system, and the control unit of the passive compensation system; the controller includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method.

[0016] The present invention has the following beneficial effects: It can accurately offset the negative impact of heave changes at the stern of mining vessels, achieving high-precision, low-energy-consumption, and stable control of umbilical cable tension under different working conditions. It effectively avoids slack impacts, reduces reliance on manual labor, and solves the technical defects of existing heave compensation technologies such as poor adaptability, rigid coupling logic, and lack of slack pre-control. It provides reliable technical support for the deployment and recovery operations of deep-sea mining vehicles, and is suitable for deep-sea mining operations in water depths of 4000-6000m. It is especially suitable for the uniform deployment or recovery operation of winches and has high engineering application value.

[0017] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0018] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the mining vehicle deployment and retrieval method based on the umbilical cable slack prediction mechanism disclosed in the embodiments of the present invention.

[0019] Figure 2 This is a schematic diagram of the overall structure of the active-passive hybrid heave compensation system disclosed in an embodiment of the present invention.

[0020] Figure 3 This is a flowchart of XGBoost heave prediction and incremental learning control based on MRU data disclosed in an embodiment of the present invention.

[0021] Figure 4 This is an example of constructing supervised learning sample data using a sliding window, as disclosed in an embodiment of the present invention.

[0022] Figure 5 This is a flowchart of the active and passive compensation execution control logic disclosed in an embodiment of the present invention.

[0023] Figure 6 This is a schematic diagram of the passive compensation system structure and threshold switching principle disclosed in the embodiments of the present invention.

[0024] Figure 7 This is a schematic diagram of the passive compensation working state disclosed in an embodiment of the present invention. Detailed Implementation

[0025] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings, but the present invention can be implemented in many different ways as defined and covered by the claims.

[0026] Example 1 This embodiment discloses a method for deploying and recovering mining vehicles based on an umbilical cable slack prediction mechanism, such as... Figure 1 As shown, the core mechanism includes the following steps S1 to S5.

[0027] Step S1: Acquire real-time heave data generated by the waves at the stern working end monitored by the attitude detection sensor, input the preprocessed real-time heave data into the heave data prediction model, and obtain the prediction result of the heave motion of the stern in the future target time.

[0028] Step S2: Based on the heave motion prediction results, combined with the winch cable length, mining vehicle speed, and current underwater depth, estimate the umbilical cable slack length D for the next moment.

[0029] Step S3: Determine whether the slack length D of the umbilical cable is still in a slack state after considering the passive compensation range corresponding to the current sea state; if yes, proceed to step S4; otherwise, proceed to step S5.

[0030] Step S4: Activate the active compensation system and perform the winch cable winding action in advance. The winding length is equal to the slack length D of the umbilical cable.

[0031] Step S5: Obtain real-time tension data of the umbilical cable and compare it with a preset tension threshold. If the real-time tension exceeds the tension threshold, start the active compensation system to perform winch cable release adjustment. When the umbilical cable tension data decreases to within the tension threshold range, switch the active compensation system to sleep mode.

[0032] Furthermore, referring to Figures 2 to 7 The steps described above in this embodiment will be further detailed below.

[0033] like Figure 2As shown, the mining vessel 1 deploys and retrieves the mining car 2. The mining car deck is equipped with an A-frame 3 for stern deployment, a winch 6 with tension and cable length monitoring functions, a passively compensated pendulum wheel 5, a motion reference unit (MRU) 7 for monitoring ship attitude and heave, and an umbilical cable 4. Specifically, the actively compensated winch 6 can actively retrieve or release the cable based on control signals to balance the tension of the umbilical cable. The passively compensated pendulum wheel 5 can oscillate within a small range when the tension changes, passively compensating for the tension of the umbilical cable 4. The mining car is equipped with an inertial measurement unit 8 and a depth sensor 9. The controllers executing steps S1 to S5 are located in the control room 10 and are connected to the attitude sensor, the control unit of the active compensation system, and the control unit of the passive compensation system.

[0034] In practical application, the active-passive hybrid heave compensation system for the deep-sea mining vehicle umbilical cable winch in this embodiment continuously activates the attitude detection sensor (MRU) deployed at the stern during the deployment or retrieval phase of the mining vehicle. This sensor monitors the vertical heave and speed of the winch's working end in real time, influenced by waves, and simultaneously collects multi-dimensional operational data such as real-time umbilical cable tension, winch cable length, mining vehicle speed, and operating water depth. The core control module first inputs the raw heave data collected by the MRU into a Kalman filter algorithm for noise reduction, filtering out high-frequency wave noise, sensor measurement interference, and other invalid data to obtain clean, continuous heave motion data. This effective heave data is continuously saved and accumulated to form historical heave data.

[0035] like Figure 3 As shown, based on the saved historical heave and sag data, the core control module learns and trains it using the XGBoost algorithm. This process is divided into two core stages: offline initial model construction and online prediction-incremental learning closed loop. The initial training and dynamic iterative updates of the model are completed using ship heave and sag data collected by the MRU. The specific process is as follows: The prediction and incremental learning process begins with the construction of the initial model. First, historical data on ship heave and sag motion is collected as the basic data source for model training. The collected historical data is preprocessed, using a Kalman filter algorithm to filter out high-frequency noise from ocean waves and sensor measurement interference, followed by data normalization to obtain clean and standardized time-series data. Subsequently, the sliding window method is used to reconstruct the one-dimensional heave and sag time-series data into supervised learning samples, that is, using previous continuous heave and sag data as features and future short-term heave and sag values ​​as labels to complete the data format conversion. The detailed process of reconstructing heave and sag prediction data training samples based on the sliding window is as follows: The MRU sensor acquires data 10 times per second, and the average heave / sag data within that second is used as the data source for training calculations. Let's assume the time... Time The heave data collected per second within a second is: , , ... The heave data for every 5 consecutive seconds is used as the feature value, and the heave data for the next 3 seconds is used as the label, thus completing the establishment of supervised learning samples.

[0036] Detailed examples are as follows Figure 4 As shown, the original data is time. Time Mean heave data per second: , , ... The sliding window method can divide the data into model input data and output data, thus obtaining supervised learning sample data, which can be used to train the prediction model.

[0037] This embodiment can use the XGBoost algorithm as the basic time series prediction model algorithm. XGBoost is a scalable gradient boosting tree algorithm disclosed by Chen Tianqi et al. in the paper "XGBoost: A Scalable Tree Boosting System" published at the 2016 ACM SIGKDD conference (specific citation information: Chen T, Guestrin C. Xgboost: A scalable tree boosting system[C] / / Proceedings of the 22nd acmsigkdd international conference on knowledge discovery and data mining. 2016:785-794.).

[0038] The XGBoost model boasts high accuracy and fast computation speed, making it well-suited for real-time prediction scenarios. It has been widely applied in industrial data prediction, time-series data analysis, and other applications. The reconstructed supervised learning samples are input into the XGBoost algorithm to train an initial XGBoost regression model. This model provides the foundation for subsequent online predictions. Once the initial model is built, the system officially enters the online prediction loop phase.

[0039] The online prediction loop runs continuously. The core control module first collects real-time heave data of the ship through the stern MRU. Based on this real-time data, a prediction feature window is constructed. That is, following the sliding window rule of the offline stage, consecutive real-time heave data from preceding periods are extracted as input features. The constructed prediction feature window is then input into an existing XGBoost regression model, which outputs the predicted heave motion of the ship in the next short period of time, providing data support for heave compensation decisions.

[0040] Set the update batch threshold to w, for example Figure 3 As shown in the process, during the online cyclic prediction phase, the system caches the latest collected MRU real-time data, continuously accumulating new sample data. After caching the latest MRU data, it checks whether the amount of cached sample data has reached the preset update batch threshold w. If it has not reached the threshold, the system continues to use the existing XGBoost for heave prediction, continues to collect MRU real-time data and perform prediction operations; if the amount of cached sample data exceeds the update batch threshold w, the incremental learning process is triggered, using all the cached new sample data to incrementally train the existing XGBoost regression model, allowing the model to learn the real-time heave motion patterns of the operating sea area and achieve dynamic optimization of the model. After the incremental training is completed, the data cache is immediately cleared, and the system returns to the starting point of the online prediction loop to continue accumulating real-time sea state heave data, waiting for the amount of cached data to exceed the threshold w to continue incremental learning and updating the XGBoost regression model. Based on the updated model, a new round of heave prediction is carried out, forming a continuous iterative closed loop of "real-time prediction - data caching - threshold judgment - incremental training - model update".

[0041] By exploring the intrinsic correlation between the heave motion of mining vessels and ocean wave changes, a heave data prediction model is constructed. Based on this model, accurate predictions of the heave motion status of the stern winch operation point can be made within a short period. Once the mining vessel arrives at the mining truck deployment area, the MRU continuously collects real-time heave data from the operation area. Simultaneously, the core control module employs an incremental learning strategy, adding newly collected heave data (after Kalman filtering and noise reduction) to the data source in real time. This dynamically updates the trained heave data prediction model, enabling it to quickly adapt to the wave characteristics and actual vessel motion patterns of the operation area, ensuring the accuracy and adaptability of the heave motion prediction results.

[0042] After training and updating the heave data prediction model, the core control module first identifies the sea state level of the current operating area based on the heave amplitude, period, and other characteristics in the heave motion prediction results. Then, it proceeds as follows... Figure 5 The logic shown executes the adaptive control of the passive compensation system, and the process is as follows: If the sea state is determined to be Class I, the wave height is 0-0.1 meters and the heave amplitude is extremely small. At this time, the effect of the ship's heave motion on the umbilical cable tension is negligible, and there is no risk of slack or impact. Therefore, the passive compensation system can be directly shut down, and the stability of the umbilical cable can be maintained by relying solely on the winch foundation tension control, thereby minimizing system energy consumption and equipment idling losses.

[0043] If the sea state is determined to be Class II, corresponding to wave heights of 0.1-0.5 meters, the ship's heave amplitude is limited, and the umbilical cable experiences only minor tension fluctuations with no risk of slack. Therefore, the accumulator of the passive compensation system is switched to a low-pressure gas cylinder via a solenoid valve to achieve a low-pressure threshold of 1. At this pressure, the system has low equivalent stiffness and small static preload, enabling passive buffering of minor tension fluctuations with low energy consumption, balancing buffering effectiveness and operational economy.

[0044] If the sea state is determined to be Class IV, corresponding to wave heights of 1.25-2.50 meters, the ship experiences severe heave and sway. The risk of instantaneous impact from the umbilical cable slackening and then re-tensioning becomes a safety hazard for the system. Therefore, the accumulator of the passive compensation system is switched to a high-pressure gas cylinder via a solenoid valve to achieve a high-pressure threshold of 2. Under this pressure, the system possesses higher initial preload and equivalent stiffness, which can prevent the umbilical cable from slackening and snapping during severe heave and swaying, thus eliminating the conditions for the generation of instantaneous impact loads to a certain extent.

[0045] If the sea state level is determined to be above level four, or if the system detects abnormal conditions such as excessive tension or heave prediction exceeding the range, a manual intervention prompt will be immediately triggered, allowing the operator to decide whether to continue the operation or to initiate the emergency recovery process for the mining vehicle, ensuring the safety of equipment and operations under extreme conditions.

[0046] In this embodiment, the passive replenishment system can be configured as follows: Figure 6 As shown, the system includes: a balance wheel 5 and a hydraulic cylinder 16, respectively fixed to the deck via a first hinge 17 and a second hinge 18. The extension and retraction of the hydraulic cylinder 16 can change the rotation amplitude of the balance wheel. An accumulator 11 is connected to the hydraulic cylinder 16, and the opening and closing of the low-pressure side solenoid valve 12 and the high-pressure side solenoid valve 13 control the connection between the low-pressure gas cylinder 15 and the high-pressure gas cylinder 14 and the accumulator 11, respectively. The pressure of the accumulator in the passive compensation system directly determines the initial preload and equivalent stiffness of the system: the higher the pressure, the greater the initial tension and the stiffer the system, resulting in stronger resistance to cable slack and impact; the lower the pressure, the smaller the static tension and the lower the energy consumption, resulting in less cable fatigue damage. In low sea states, the core requirements are low energy consumption and long lifespan. Therefore, a low-pressure threshold is adopted to achieve small-scale buffering with low stiffness and low static tension, avoiding long-term heavy load fatigue of the cable. In high sea states, the core requirements are shock resistance and safety. Therefore, a high-pressure threshold is adopted to ensure the umbilical cable is tensioned throughout the entire process with high pretension and high stiffness, fundamentally avoiding rapid relaxation and tensioning of the umbilical cable, avoiding the fatal risk of instantaneous hard impact, and significantly improving the overall safety margin of the system.

[0047] While completing sea state identification and passive compensation threshold adjustment, the core control module inputs the preprocessed real-time heave data into the updated heave data prediction model to obtain the prediction results of the heave motion of the stern in the near future. Then, combined with key parameters such as the current winch cable length and the speed of the mining vehicle, the slack prediction model is used to calculate and analyze whether the umbilical cable will enter a slack state.

[0048] Reference Figure 7 Let L0 be the length of the umbilical cable at the moment the mining vehicle enters the water, and L1 be the current length of the umbilical cable. Let time be t, the vehicle's axial velocity obtained by the mining vehicle's IMU (Inertial Measurement Unit) be v (positive if the vehicle is moving downwards, negative if the vehicle is moving upwards), the distance the ship will descend at time t predicted by the model be L3, the deformation ratio of the umbilical cable with its length be k, the current underwater depth of the mining vehicle be H, and the relaxation calculation formula for the next t seconds is: .

[0049] In the above calculation formula, the total length of the newly laid umbilical cable is obtained by subtracting the cable car length L0 at the initial entry point into the water from the total length of the laid cable. Considering that the weight of the underwater umbilical cable increases with the length laid, and the tensile deformation also increases, the cable length L1 recorded by the winch is not equal to the actual length of the underwater umbilical cable. Therefore, k×(L1-L0) represents the elongation due to deformation of the umbilical cable, and adding (L1-L0) gives the total length of the underwater umbilical cable. Meanwhile, based on... The estimated depth of the vehicle entering the water at time t can be obtained.

[0050] After obtaining the umbilical cable relaxation displacement D based on the above formula, the relaxation displacement D is compared with the passive compensation displacement range L2. If D is greater than L2, it means that the passive compensation cannot make up for the displacement at time t in the future, that is, the length of the umbilical cable being lowered just exceeds the "effective length to maintain tension", and it is about to enter the relaxation state.

[0051] If the prediction indicates that the umbilical cable is about to become slack, the core control module immediately activates the active compensation system based on the calculated estimated slack length of the umbilical cable. This drives the active drive unit to perform the winch cable retrieval action, matching the retrieval length to the estimated slack length, thus eliminating the risk of umbilical cable slack in advance. If the prediction indicates that the umbilical cable will not become slack, the core control module further reads the real-time tension data of the umbilical cable and compares it with a preset tension threshold. If the real-time tension exceeds the set threshold, the active compensation system is immediately activated to perform the winch cable release adjustment action, reducing the umbilical cable tension to within the threshold range. If the real-time tension does not exceed the set threshold, the active compensation system is switched to a dormant state, relying solely on the passive compensation system, which has been configured with parameters according to sea state levels, to maintain the basic tension buffer. The winch can maintain its current normal deployment or retrieval speed.

[0052] In summary, the mining vehicle deployment and retrieval method based on the umbilical cable slack prediction mechanism disclosed in this embodiment has at least the following beneficial effects: (i) It realizes the passive compensation working threshold of sea state level allocation, abandons the existing fixed threshold mode, and dynamically allocates the compensation working pressure according to the actual working conditions, so as to meet the heavy-load and high-precision operation requirements of deep-sea mining vehicles, and takes into account the compensation performance and energy consumption economy.

[0053] (ii) A slack prediction model for umbilical cables based on the stern motion trend was established. This model can predict slack risks in advance and generate active pre-tightening commands, thereby avoiding instantaneous impact loads caused by slack followed by re-tightening, reducing the risk of umbilical cable damage, and extending its service life.

[0054] (III) The entire process of heave compensation has been automated. From sensor data acquisition and heave prediction to compensation execution, no manual intervention is required. Only basic operation parameters need to be set manually, which greatly reduces the difficulty of operation.

[0055] (iv) The system adopts a modular integrated design. Each module can be directly integrated into the existing umbilical cable winch equipment of the deep-sea mining vessel through the communication interface. There is no need to make major modifications to the mother ship deck. It occupies little space, has low modification cost, and has good versatility and engineering adaptability.

[0056] (V) It can accurately offset the negative impact of the heave changes at the stern of the mining vessel, achieve high-precision, low-energy-consumption stable control of the umbilical cable tension under different working conditions, effectively avoid slack impact, reduce manual dependence, and solve the technical defects of poor adaptability, rigid coupling logic and lack of slack pre-control of existing heave compensation technology. It provides reliable technical support for the deployment and recovery of deep-sea mining vehicles, and is suitable for deep-sea mining operations in water depths of 4000-6000m. It is especially suitable for the uniform deployment or recovery operation of winches and has high engineering application value.

[0057] Example 2 This embodiment discloses a mining truck deployment and retrieval system based on an umbilical cable slack prediction mechanism, comprising: an A-frame for stern deployment mounted on the deck of the mining truck; a winch with tension and cable length monitoring functions; a pendulum wheel for passive compensation; an attitude sensor for monitoring real-time heave data generated by waves at the stern working end; and a connection to the umbilical cable; the active compensation winch can actively retrieve or release the cable based on control signals to balance the tension of the umbilical cable; the passive compensation pendulum wheel can swing within a small range when the tension changes, passively compensating for the tension of the umbilical cable; the mining truck is equipped with an inertial navigation IMU and a depth sensor; a controller located in the control room establishes signal connections with the attitude sensor, the control unit of the active compensation system, and the control unit of the passive compensation system; the controller includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method.

[0058] The underlying logic and functionality of this embodiment are the same as those of Embodiment 1 above, and will not be repeated here.

[0059] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for deploying and recovering mining vehicles based on an umbilical cable slack prediction mechanism, characterized in that, include: Step S1: Acquire real-time heave data generated by the waves at the stern working end monitored by the attitude detection sensor, input the preprocessed real-time heave data into the heave data prediction model, and obtain the prediction result of the heave motion of the stern in the future target time. Step S2: Based on the heave motion prediction results, combined with the winch cable length, mining vehicle speed and current underwater depth, predict the umbilical cable slack length D at the next moment. Step S3: Determine whether the slack length D of the umbilical cable is still in a slack state after considering the passive compensation range corresponding to the current sea state; If yes, proceed to step S4; otherwise, proceed to step S5. Step S4: Activate the active compensation system and perform the winch cable retrieval action in advance. The retrieved cable length is equal to the slack length D of the umbilical cable. Step S5: Obtain real-time tension data of the umbilical cable and compare it with a preset tension threshold. If the real-time tension exceeds the tension threshold, start the active compensation system to perform winch cable release adjustment. When the umbilical cable tension data decreases to within the tension threshold range, switch the active compensation system to sleep mode.

2. The mining vehicle deployment and retrieval method based on umbilical cable slack prediction mechanism according to claim 1, characterized in that, The steps preceding step S1 also include: Training samples for the heave data prediction model are reconstructed using a sliding window based on historical heave data from the attitude detection sensor in the operating sea area.

3. The method for deploying and recovering mining vehicles based on an umbilical cable slack prediction mechanism according to claim 2, characterized in that, The heave data prediction model uses the XGBoost algorithm for initial model construction and incremental learning training. The attitude detection sensor collects data 10 times per second. The training samples for reconstructing the heave data prediction model through a sliding window are specifically as follows: the average heave data per second is taken as the data source for training operations; the heave data for every 5 consecutive seconds is taken as the feature value; and the heave data for the next 3 seconds is taken as the label, so as to construct supervised learning samples.

4. The method for deploying and recovering mining vehicles based on an umbilical cable slack prediction mechanism according to claim 2, characterized in that, The sea state is divided into Level 1, Level 2, Level 4 and above; the passive compensation system is in the off state when the sea state is Level 1, and the corresponding passive compensation range is 0; the passive compensation system is activated when the sea state is at least Level 2.

5. The method for deploying and recovering mining vehicles based on an umbilical cable slack prediction mechanism according to claim 4, characterized in that, The passive compensation system comprises: a balance wheel and a hydraulic cylinder fixed to the deck by a hinge; the extension and retraction of the hydraulic cylinder can change the rotation amplitude of the balance wheel; one end of the accumulator is connected to the hydraulic cylinder, and the other end is connected to the low-pressure gas cylinder and the high-pressure gas cylinder respectively through the opening and closing of the low-pressure side solenoid valve and the high-pressure side solenoid valve. The method further includes: when the sea state is identified as level 2 based on a series of continuous real-time heave data, opening the low-pressure side solenoid valve and closing the high-pressure side solenoid valve; when the sea state is identified as at least level 4 based on a series of continuous real-time heave data, closing the low-pressure side solenoid valve and opening the high-pressure side solenoid valve.

6. The method for deploying and recovering mining vehicles based on an umbilical cable slack prediction mechanism according to any one of claims 1 to 4, characterized in that, The formula for calculating the slack length D of the umbilical cord is: Where L0 is the length of the umbilical cable at the moment the mining vehicle enters the water, L1 is the current length of the umbilical cable, k is the deformation ratio of the umbilical cable with length, L3 is the distance the ship will descend at the next moment as predicted by the heave data prediction model, t is the time interval between adjacent calculation frequencies, H is the underwater depth of the mining vehicle at the current moment, and v is the vehicle's azimuth velocity obtained by the mining vehicle's inertial measurement unit.

7. A mining vehicle deployment and retrieval system based on an umbilical cable slack prediction mechanism, characterized in that, include: The mining vehicle deck is equipped with an A-frame for stern deployment, a winch with tension and cable length monitoring functions, a pendulum wheel for passive compensation, an attitude sensor for monitoring real-time heave data generated by waves at the stern working end, and a connection to the umbilical cable. The active compensation winch can actively retrieve or release the cable based on control signals to balance the tension of the umbilical cable. The passive compensation pendulum wheel can swing within a small range when the tension changes, passively compensating for the tension of the umbilical cable. The mining vehicle is equipped with an inertial measurement unit and a depth sensor. The controller located in the control room establishes signal connections with the attitude sensor, the control unit of the active compensation system, and the control unit of the passive compensation system. The controller includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described in any one of claims 1 to 6.

Images