Method for constructing forecasting and early warning model for accidental excessive rolling motion

Abstract

The invention belongs to the technical field of ship rolling motion early warning, and discloses a method for constructing a forecasting and early warning model for accidental excessive rolling motion, and the model comprises a data collection module, a data processing module, a state estimation module and a forecasting and early warning module. The construction method of the model comprises the following steps: collecting ship motion data and sea condition data in real time based on a ship motion sensor; preprocessing the collected ship motion data and sea condition data, and constructing an input module of the ship intelligent sensing system; establishing a ship motion adaptive response model under all working conditions and all sea conditions; and a forecasting and early warning mechanism of accidental excessive rolling motion is established. According to the construction method of the forecasting and early warning model for the accidental excessive rolling motion, the forecasting and early warning model for the accidental excessive rolling motion is established, the excessive rolling of the ship is forecasted and early warned in advance, the ship navigation safety under the severe sea condition is improved, timely decision support is provided for sailors, and the safe navigation of the ship is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of ship roll motion early warning, and particularly relates to a method for constructing a prediction and early warning model for accidental excessive roll motion. Background Art

[0002] In the magnificent ocean century, ships, as the main transportation carriers, are widely used in various aspects such as developing the ocean, utilizing the ocean, protecting the ocean, and controlling the ocean. Ships are an important support for maritime transportation and the ocean economy. How to improve the navigation safety of ships has always been a research hotspot and difficulty for scholars at home and abroad.

[0003] When sailing on the unpredictable sea, maritime accidents occur frequently. Among them, ship losses caused by excessive rolling under severe sea conditions are the most common, usually resulting in huge property losses, serious casualties, and irreparable marine environmental pollution. The factors leading to excessive rolling of ships include resonance rolling, parametric rolling, pure loss of stability, dead ship stability, surfing / sideslip, and excessive acceleration, etc. Among them, beam sea resonance rolling and parametric rolling are the most serious.

[0004] Traditional ship roll reduction devices include fin stabilizers, anti-rolling tanks, anti-rolling gyros, and bilge keels, etc. However, except for bilge keels becoming standard equipment, the application of other roll reduction devices is restricted by many factors. Taking fin stabilizers as an example, as a device with better roll reduction effect, due to reasons such as large occupied space, high installation cost, high operation and maintenance cost, and large towing resistance, its utilization rate on ordinary merchant ships is relatively low. In contrast, rudder roll reduction can be carried out based on the existing course rudder equipment, and it has both economy and safety. The research on rudder roll reduction has been going on for more than half a century, and the application of roll reduction rudder products is only limited to ships in some countries and has not been widely promoted. According to the follow-up research on the theory and products of rudder roll reduction, there are significant dual uncertainties in rudder roll reduction control. On the one hand, the superposition of initial steering inboard inclination, turning outboard inclination, and external disturbance-induced rolling may induce operational excessive rolling problems. Especially when facing situations such as beam sea resonance rolling, parametric rolling, and surfing sideslip, it is easy to make roll reduction become roll augmentation, resulting in roll reduction safety uncertainty. On the other hand, model perturbation caused by changes in navigation conditions, external random disturbances such as wind, waves, and currents, as well as constraints such as the nonlinearity, rapidity, and loss of the rudder machine, affect the robustness and self-adaptability of the control algorithm, restricting the ship rudder roll reduction control effect and resulting in serious roll reduction performance uncertainty problems.

[0005] However, at present, the shipping industry is accustomed to the large-amplitude rolling of ships under strong winds and waves, and generally believes that based on reasonable cargo loading, ship strength meeting regulatory requirements, and skilled navigation skills, etc., it can all be overcome. However, numerous marine accidents have proved that the excessive rolling that poses a safety hazard is accidental and cannot be ignored.

[0006] The occasional excessive rolling in severe sea conditions is a problem that must be faced in the development of ship intelligence and unmanned operation. Therefore, the present invention proposes a method for constructing a prediction and early warning model for occasional excessive rolling motion. Summary of the Invention

[0007] The object of the present invention is to provide a method for constructing a prediction and early warning model for occasional excessive rolling motion, which solves the safety problems such as the sharp reduction or even loss of ship stability, damage to the hull structure strength, and failure of cargo lashing caused by occasional excessive rolling in severe sea conditions. The active safety theory is applied to the control of the ship's fin stabilizer. When the ship faces the danger caused by excessive rolling, the excessive rolling of the ship is predicted and warned in advance to avoid the occurrence of marine accidents, prevent problems before they happen, effectively improve the ship navigation safety in severe sea conditions, provide timely decision-making support for the crew, and ensure the safe navigation of the ship.

[0008] To achieve the above object, the present invention provides a method for constructing a prediction and early warning model for occasional excessive rolling motion. The model includes a data acquisition module, a data processing module, a state estimation module, and a prediction and early warning module. Among them, the method for constructing the model includes the following steps:

[0009] Step S1: Based on ship motion sensors, ship motion data and sea condition data are collected in real time.

[0010] Step S2: The collected ship motion data and sea condition data are preprocessed to construct the input module of the ship intelligent perception system.

[0011] Step S3: An adaptive response model of ship motion under all working conditions and all sea conditions is established.

[0012] Step S4: A prediction and early warning mechanism for occasional excessive rolling motion is established.

[0013] Preferably, in step S1, the ship motion data includes different loading states such as light load, half load, and full load, and different speed states such as full speed, half speed, slow speed, and micro speed; the sea condition data includes no waves, light waves, medium waves, and large waves.

[0014] Preferably, in step S2, the collected ship motion data and sea condition data are preprocessed to construct the input module of the ship intelligent perception system. The specific process is as follows:

[0015] Step S21: The collected ship motion data and sea condition data are subjected to non-linear filtering processing by using cubature Kalman filtering, multi-source heterogeneous information fusion, and iterative intelligent filtering algorithms.

[0016] Step S22: The sample data is quantitatively and qualitatively compared with the output of the non-linear motion model to verify the reliability of the filtering module and the accuracy of the data.

[0017] Step S23: Based on the above steps, construct the data input module of the ship intelligent perception system.

[0018] Preferably, in step S3, establish a ship motion adaptive response model under all working conditions and all sea conditions, and the specific process is as follows:

[0019] Step S31: For the yaw-roll coupling mechanism under the initial stability and large-angle stability conditions of a ship with rudder and propeller configuration, combine the Isidori nonlinear stability theory and safety pool analysis to conduct simulation tests on ship motion data and data analysis;

[0020] Step S32: Adopt the idea similar to Nomoto to establish a ship roll response model to rudder, and analyze the structure and physical meaning of parameters of the ship roll response model according to the typical motion links of initial heel, steady outboard heel and periodic roll;

[0021] Step S33: Combine multiple parameter identification algorithms such as nonlinear GA, least squares method, support vector machine and empirical parameter identification to conduct parameter identification of the ship roll response model to rudder;

[0022] Step S34: Conduct qualitative and quantitative analysis and research on the perturbation range of ship motion response model parameters under different loadings, speeds and sea conditions, and establish a ship motion adaptive response characteristic model under all working conditions and all sea conditions.

[0023] Preferably, in step S34, establish a ship motion adaptive response characteristic model under all working conditions and all sea conditions, and the specific process is as follows:

[0024] Step S341: Take the loading state as the first weight, and establish ship motion response models under the conditions of no load, half load and full load;

[0025] Step S342: Take the speed as the second weight and establish ship motion response models under the conditions of full speed, half speed, slow speed and micro speed for a certain loading state;

[0026] Step S343: Take the sea condition as the third weight and establish ship motion response models under the sea conditions of no waves, small waves, medium waves and large waves for a certain speed state;

[0027] Step S345: Use neural network deep learning, fuzzy synthesis and multi-model switching methods to construct a ship motion adaptive response characteristic model under all working conditions and all sea conditions.

[0028] Preferably, in step S4, establish a prediction and early warning mechanism for accidental excessive roll motion, and the specific process is as follows:

[0029] Step S41: Based on the ship motion sensor data acquisition and the ship motion adaptive response characteristic model, establish a ship navigation safety perception system;

[0030] Step S42: Relying on the distributed sensors for real ship motion acquisition, use algorithms such as limit filtering and nonlinear Kalman filtering to perform singular point elimination and interference filtering, and extract the motion data of roll, yaw, sway, and surge;

[0031] Step S43: Select a suitable response model according to the loading condition, speed, and sea state, and establish a prediction model for accidental excessive roll motion;

[0032] Step S44: Based on the IMO ship stability criteria requirements, determine the safety standards for excessive roll and its risk identification principles, clarify the superposition mechanisms such as steering heel, interference roll, and turning heel, and form a dangerous roll warning standard; respectively establish warning models for accidental excessive roll motion from aspects of beam sea resonance rolling, parametric rolling, pure stability loss, dead ship stability, surfing / broaching, and excessive acceleration stability failure modes, and form an active safety warning mechanism for ship excessive roll motion.

[0033] Therefore, the present invention adopts the above-mentioned method for constructing a prediction and warning model for accidental excessive roll motion, solves the safety problems such as the sharp reduction or even loss of ship stability, damage to the hull structure strength, and failure of cargo lashing caused by accidental excessive roll in severe sea conditions, applies the active safety theory to the control of ship anti-rolling fins, predicts and warns the excessive roll of the ship in advance when the ship faces the danger caused by excessive roll, avoids the occurrence of marine accidents, takes precautions, effectively improves the ship navigation safety in severe sea conditions, provides timely decision-making support for the crew, and ensures the safe navigation of the ship.

[0034] The following further describes the technical solution of the present invention in detail through the drawings and embodiments. Description of the Drawings

[0035] Figure 1 is a module diagram of a prediction and warning model for accidental excessive roll motion of the present invention;

[0036] Figure 2 is a flowchart of the method for constructing a prediction and warning model for accidental excessive roll motion of the present invention;

[0037] Figure 3 is the construction process of the ship motion adaptive response characteristic model under all working conditions and all sea conditions of the present invention;

[0038] Figure 4 is the technical route of the method for constructing a prediction and warning model for accidental excessive roll motion of the present invention. Detailed Embodiments

[0039] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0040] As Figure 1 shown, a prediction and early warning model for occasional excessive rolling motion includes a data acquisition module, a data processing module, a state estimation module, and a prediction and early warning module;

[0041] As Figure 2 shown, a construction method of a prediction and early warning model for occasional excessive rolling motion includes the following steps:

[0042] Step S1: Based on ship motion sensors, ship motion data and sea condition data are collected in real time.

[0043] The ship motion data includes different loading states such as light load, half load, and full load, and different speed states such as full speed, half speed, slow speed, and micro speed.

[0044] The sea condition data includes: no waves, light waves, moderate waves, and large waves, etc.

[0045] Step S2: Preprocess the collected ship motion data and sea condition data to construct the input module of the ship intelligent perception system.

[0046] Step S21: Use algorithms such as cubature Kalman filter, multi-source heterogeneous information fusion, and iterative intelligent filter to perform non-linear filtering on the collected ship motion data and sea condition data.

[0047] Step S22: Quantitatively and qualitatively compare the sample data with the output of the non-linear motion model to verify the reliability of the filtering module and the accuracy of the data.

[0048] Step S23: Based on the above steps, construct the data input module of the ship intelligent perception system.

[0049] Step S3: Establish an adaptive response model of ship motion under all working conditions and all sea conditions.

[0050] Step S31: For the yaw-roll coupling mechanism under the initial stability and large-angle stability conditions of a ship with rudder and propeller configuration, combine the Isidori non-linear stability theory and safety pool analysis to conduct simulation tests and data analysis on the ship motion data.

[0051] Step S32: Adopt the idea similar to Nomoto to establish a ship roll response model to the rudder, and analyze the physical meaning of the structure and parameters of the ship roll response model according to typical motion links such as initial heel, stable heel, and periodic roll.

[0052] Step S33: Identify the parameters of the ship's roll-to-steering response model by combining multiple parameter identification algorithms such as non-linear GA, least squares method, support vector machine, and empirical parameter identification.

[0053] Step S34: Conduct qualitative and quantitative analysis on the perturbation ranges of the parameters of the ship motion response model under different loading conditions, speeds, and sea states, and establish an adaptive response characteristic model for ship motion under all working conditions and all sea states, as Figure 3 shown.

[0054] Step S341: Taking the loading state as the first weight, establish ship motion response models for states such as light load, half load, and full load.

[0055] Step S342: Taking the speed as the second weight, establish ship motion response models for ship speeds such as full speed, half speed, slow speed, and micro speed under a certain loading state.

[0056] Step S343: Taking the sea state as the third weight, establish ship motion response models for sea states such as calm sea, small waves, moderate waves, and large waves under a certain speed state.

[0057] Step S345: Use methods such as neural network deep learning, fuzzy synthesis, and multi-model switching to construct an adaptive response characteristic model for ship motion under all working conditions and all sea states.

[0058] Step S4: Establish a prediction and early warning mechanism for accidental excessive rolling motion, as Figure 4 shown.

[0059] Step S41: Based on ship motion sensor data collection and the ship motion adaptive response characteristic model, establish a ship navigation safety perception system.

[0060] Step S42: Relying on the distributed sensor collection of actual ship motion, use algorithms such as limit filtering and non-linear Kalman filtering to perform singular point removal and interference filtering, and extract motion data such as roll, yaw, sway, and surge.

[0061] Step S43: Select an appropriate response model according to the loading state, speed, and sea state, and establish a prediction model for accidental excessive rolling motion.

[0062] Step S44: Based on the IMO ship stability criteria requirements, etc., determine the safety standards for excessive rolling and its risk identification principles, clarify the superposition mechanisms such as steering heel, interference roll, and turning heel, and form a dangerous roll early warning standard; establish early warning models for accidental excessive rolling motion from stability failure modes such as beam sea resonance rolling, parametric rolling, pure stability loss, dead ship stability, surfing / broaching, and excessive acceleration, and form an active safety early warning mechanism for ship excessive rolling motion.

[0063] Therefore, by adopting the above method for constructing a prediction and early warning model for accidental excessive rolling motion, the present invention solves the safety problems such as the sharp reduction or even loss of ship stability, damage to the hull structure strength, and failure of cargo lashing caused by accidental excessive rolling in severe sea conditions. It applies the active safety theory to the control of ship anti-rolling fins. When the ship faces the danger caused by excessive rolling, it can predict and early warn the excessive rolling of the ship in advance, avoid the occurrence of marine accidents, take precautions, effectively improve the ship navigation safety in severe sea conditions, provide timely decision-making support for the crew, and ensure the safe navigation of the ship.

[0064] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for constructing a forecast and warning model for occasional excessive rolling motion, characterized in that: The model includes a data acquisition module, a data processing module, a state estimation module and a forecast and warning module; wherein the construction method of the model includes the following steps: Step S1, based on the ship motion sensor, real-time collection of ship motion data and sea condition data; Step S2: pre-processing the collected ship motion data and sea condition data to construct an input module of the ship intelligent perception system; Step S3, establishing a ship motion adaptive response model under all working conditions and all sea conditions; Step S4: establishing a forecast and warning mechanism for occasional excessive rolling motion.

2. The method for constructing a forecast and warning model for occasional excessive rolling motion according to claim 1, characterized in that: In step S1, the ship motion data includes: different loading states of empty, half-loaded and fully loaded, and different speed states of full speed, half speed, slow speed and slow speed; the sea condition data includes: no waves, light waves, medium waves and big waves.

3. The method for constructing a forecast and warning model for occasional excessive rolling motion according to claim 1, characterized in that: In step S2, the collected ship motion data and sea condition data are preprocessed to construct an input module of the ship intelligent perception system. The specific process is as follows: Step S21, using volumetric Kalman filtering, multi-source heterogeneous information fusion and iterative intelligent filtering algorithm to perform nonlinear filtering processing on the collected ship motion data and sea state data; Step S22: perform quantitative and qualitative comparison between the sample data and the output of the nonlinear motion model to verify the reliability of the filtering module and the accuracy of the data; Step S23: Based on the above steps, construct a data input module of the ship intelligent perception system.

4. The method for constructing a forecast and warning model for occasional excessive rolling motion according to claim 1, characterized in that: In step S3, an adaptive response model of ship motion under all working conditions and all sea conditions is established. The specific process is as follows: Step S31, for the initial stability of the rudder-equipped ship and the bow-roll coupling mechanism under the large inclination angle stability conditions, combining Isidori nonlinear stability theory and safety pool analysis, simulate the ship motion data and perform data analysis; Step S32: adopting a Nomoto-like idea to establish a ship roll response model to the rudder, and analyzing the structure of the ship roll response model and the physical meaning of the parameters according to the typical motion links of initial inclination, stable outclination and periodic rolling; Step S33, using a combination of multiple parameter identification algorithms to perform parameter identification of the ship rolling response model to the rudder; Step S34, conduct qualitative and quantitative analysis on the perturbation range of the ship motion response model parameters under different loadings, speeds and sea conditions, and establish an adaptive response characteristic model of ship motion under all working conditions and all sea conditions.

5. The method for constructing a forecast and warning model for occasional excessive rolling motion according to claim 4, characterized in that: In step S34, an adaptive response characteristic model of ship motion under all working conditions and all sea conditions is established. The specific process is as follows: Step S341, taking the loading state as the first weight, respectively establishing a ship motion response model under the load states of no load, half load and full load; Step S342, establishing a ship motion response model under full speed, half speed, slow speed and slow speed ship speed under a certain loading state with the ship speed as the second weight; Step S343, using the sea condition as the third weight, establishing a ship motion response model under no wave, small wave, medium wave and large wave sea conditions at a certain speed state; Step S345: Use neural network deep learning, fuzzy synthesis and multi-model switching methods to construct an adaptive response characteristic model of ship motion under all working conditions and all sea conditions.

6. The method for constructing a forecast and warning model for occasional excessive rolling motion according to claim 1, characterized in that: In step S4, a forecast and warning mechanism for occasional excessive rolling motion is established. The specific process is as follows: Step S41: Establishing a ship navigation safety perception system based on ship motion sensor data collection and ship motion adaptive response characteristic model; Step S42, relying on the distributed sensor collection of real ship motion, using limiting filtering and nonlinear Kalman filtering algorithms, etc., to remove singular points and perform interference filtering processing to extract the motion data of roll, pitch, sway and pitch; Step S43, selecting a suitable response model according to the loading state, ship speed and sea conditions, and establishing a prediction model for occasional excessive rolling motion; Step S44: Based on the IMO ship stability criteria, the safety standards for excessive roll and its risk identification principles are determined, the superposition mechanisms of steering inclination, interference roll and gyration outclination are explained, and a dangerous roll warning standard is formed; from the transverse wave harmonic roll, parameter roll, pure stability loss, paralyzed ship stability, wave riding / rolling and excessive acceleration stability failure modes, a warning model for occasional excessive roll motion is established to form an active safety warning mechanism for excessive roll motion of the ship.

Images