A bridge grab crane operation process optimization method based on time series data

Abstract

The application provides a bridge grab crane operation process optimization method based on time series data, which comprises the following steps: collecting real-time time series data in the operation process of the bridge grab crane; constructing a hidden Markov model-based operation process state space model; performing state recognition and time series segmentation by using a forward algorithm and a Viterbi algorithm; constructing a comprehensive efficiency evaluation model comprising time efficiency, space efficiency and energy consumption efficiency; estimating model parameters based on a Baum-Welch algorithm, and performing multi-objective optimization on the operation cycle, energy consumption and operation fluctuation in combination with a genetic algorithm. The application can accurately identify the operation stage, realize multi-dimensional efficiency evaluation and dynamic optimization, and experimental results show that the operation efficiency is improved by 12.4% and the energy consumption is reduced by 13.6%, so the application can be widely applied to bulk material handling scenes such as ports, power plants and metallurgy.

Description

Technical Field

[0001] This invention belongs to the field of industrial automation and intelligent control technology, specifically relating to a method for optimizing the operation process of a bridge grab crane based on time series data. Background Technology

[0002] Bridge grab cranes are core equipment in bulk material handling operations in ports, docks, power plants, and metallurgical enterprises, and are widely used for handling bulk materials such as coal, ore, and grain. Their typical operating process consists of four stages: "grab, release, and unload." The grabbing process can be further divided into three continuous actions: "lowering, closing, and lifting," while the unloading process is divided into three continuous actions: "lowering, opening, and lifting."

[0003] Traditional bridge grab cranes rely mainly on manual operation or simple sequential control logic for operation control, which presents the following technical problems:

[0004] 1. Rough identification of operation phases: The existing system has difficulty in accurately distinguishing the sub-phases such as descent, closure, and lifting in the grasping process, which makes it impossible to optimize control parameters in a targeted manner.

[0005] 2. Limited efficiency evaluation dimensions: Existing methods typically focus only on work cycle time, neglecting multi-dimensional efficiency indicators such as energy consumption and operational stability.

[0006] 3. Lack of mathematical modeling support: The work process lacks a unified state-space model, making it difficult to achieve data-driven fine-grained control and optimization.

[0007] 4. Insufficient optimization capability: Most existing control strategies are based on static preset parameters and cannot be dynamically optimized and adjusted according to real-time data.

[0008] Therefore, there is an urgent need for a technical solution that can accurately model, identify the status, and optimize the efficiency of bridge grab crane operation based on real-time time series data. Summary of the Invention

[0009] The present invention aims to provide a method for optimizing the operation process of a bridge grab crane based on time series data, in order to solve the problems of inaccurate identification of operation stages, single efficiency evaluation, lack of mathematical modeling support and insufficient optimization capabilities in the existing technology.

[0010] To achieve the objectives of this invention, the technical solution adopted is as follows:

[0011] A method for optimizing the operation process of a bridge grab crane based on time series data, the method comprising:

[0012] S1. Collect real-time time series data of the bridge grab crane during operation;

[0013] S2. Based on the real-time time series data and hidden Markov model in step S1, construct the state space model of the work process. ,in For a set of states, For the observation set, Here is the state transition matrix. For the observation probability matrix, This is the initial state probability vector;

[0014] S3. Based on real-time time series data, the forward algorithm and Viterbi algorithm are used to identify the status of the work process and segment the time series, and output the status sequence of the work stage.

[0015] S4. Based on the identified state sequence, calculate the time efficiency index, spatial efficiency index and energy consumption efficiency index respectively, and construct a comprehensive efficiency evaluation model.

[0016] S5. Estimate the parameters of the hidden Markov model based on the Baum-Welch algorithm, and combine the genetic algorithm to perform multi-objective optimization of the operation cycle, energy consumption and operation fluctuation, and output the optimized operation control parameters.

[0017] Furthermore, the real-time time series data in step S1 includes position coordinates, height information, load information, rope tension, grab torque, and motor current.

[0018] Furthermore, in step S2:

[0019] State set The states are, in sequence, preparation state, outbound state, grab-and-release state, grab-and-close state, grab-and-lift state, return state, release-and-release state, release-and-open state, and release-and-lift state.

[0020] Observation set , This includes location coordinates, altitude information, load information, rope tension, grab torque, and motor current;

[0021] State transition matrix ,in ;

[0022] Observation probability matrix ,in ;

[0023] Initial state probability vector .

[0024] Furthermore, the comprehensive efficiency evaluation model in step S4 is expressed as follows:

[0025] ;

[0026] in, Indicates time efficiency. Indicates space efficiency. Indicates energy efficiency. , , Let represent the weighting coefficients, and + + =1.

[0027] Furthermore, the multi-objective optimization of the operation cycle, energy consumption, and operational fluctuations in step S5 is expressed as follows:

[0028] ;

[0029] in: This indicates minimizing the job cycle. This represents minimizing total energy consumption. This indicates minimizing operational fluctuations.

[0030] Furthermore, the fitness function of the genetic algorithm in step S5 is as follows:

[0031] ;

[0032] in: .

[0033] Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:

[0034] 1. A work process state space containing refined states was constructed based on the Hidden Markov Model. The state recognition was performed by combining the forward algorithm and the Viterbi algorithm, and the average recognition accuracy reached 96.8%. It can accurately distinguish the sub-stages such as descent, closing / opening, and lifting in the grasping and unloading process.

[0035] 2. By constructing a comprehensive evaluation model that includes time efficiency, space efficiency, and energy efficiency, the limitations of traditional methods that only focus on a single indicator are overcome, providing a scientific basis for multi-objective optimization.

[0036] 3. The Baum-Welch algorithm is used for parameter estimation, and combined with the genetic algorithm to achieve multi-objective optimization. It can dynamically adjust the control parameters according to real-time data to adapt to different working conditions.

[0037] 4. After adopting the method of the present invention, the operating efficiency of the bridge grab crane is increased by 12.4% and the energy consumption is reduced by 13.6%, which has good economic benefits and promotion prospects.

[0038] 5. The state-space modeling method and optimization framework proposed in this invention can be extended to more complex application scenarios such as multi-device collaborative operation and digital twin systems. Attached Figure Description

[0039] Figure 1 This is a flowchart of the method of the present invention;

[0040] Figure 2 This is a state transition diagram in an embodiment of the present invention;

[0041] Figure 3 This is a time-series state distribution diagram in an embodiment of the present invention;

[0042] Figure 4 This is a comparison chart of the time distribution before and after efficiency optimization in an embodiment of the present invention. Detailed Implementation

[0043] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0044] Example 1:

[0045] This embodiment provides a method for optimizing the operation process of a bridge grab crane based on time series data, such as... Figure 1 As shown, the method includes:

[0046] S1. Collect real-time time series data of the bridge grab crane during operation, including position coordinates, height information, load information, rope tension, grab torque and motor current.

[0047] S2. Based on the real-time time series data and hidden Markov model in step S1, construct the state space model of the work process. ,in:

[0048] State set The states are, in sequence, preparation state, outbound state, grab-and-release state, grab-and-close state, grab-and-lift state, return state, release-and-release state, release-and-open state, and release-and-lift state.

[0049] Observation set , This includes location coordinates, altitude information, load information, rope tension, grab torque, and motor current;

[0050] State transition matrix ,in ;

[0051] Observation probability matrix ,in ;

[0052] Initial state probability vector .

[0053] S3. Based on real-time time series data, the forward algorithm and Viterbi algorithm are used to identify the status of the work process and segment the time series, and output the status sequence of the work stage.

[0054] S4. Based on the identified state sequence, calculate the time efficiency index, spatial efficiency index and energy consumption efficiency index respectively, and construct a comprehensive efficiency evaluation model.

[0055] The comprehensive efficiency evaluation model is expressed as follows:

[0056] ;

[0057] in, Indicates time efficiency. Indicates space efficiency. Indicates energy efficiency. , , Let represent the weighting coefficients, and + + =1.

[0058] S5. Estimate the parameters of the hidden Markov model based on the Baum-Welch algorithm, and combine the genetic algorithm to perform multi-objective optimization of the operation cycle, energy consumption and operation fluctuation, and output the optimized operation control parameters.

[0059] The multi-objective optimization of work cycle, energy consumption, and operational fluctuations is expressed as follows:

[0060] ;

[0061] in: This indicates minimizing the job cycle. This represents minimizing total energy consumption. This indicates minimizing operational fluctuations;

[0062] The fitness function of the genetic algorithm is as follows:

[0063] ;

[0064] in: .

[0065] Example 2:

[0066] Based on Example 1, the following detailed optimizations are performed, such as... Figure 2 , 3 As shown in Figure 4:

[0067] 1. Establishment of Mathematical Model

[0068] 1.1 System State-Space Model

[0069] Let the state space of the bridge grab crane system be... Each state represents a different stage of the operation:

[0070] Ready State

[0071] Outbound State

[0072] Grabbing-Lowering State

[0073] Grabbing-Closing State

[0074] Grabbing-Raising State

[0075] Return State

[0076] Releasing-Lowering State

[0077] Releasing-Opening State

[0078] Releasing-Raising State

[0079] The state transitions of a system can be described using Markov chains:

[0080] ;

[0081] in, Indicates from state Transition to state The probability of.

[0082] 1.2 Time Series Observation Model

[0083] Let the time series observation vector be... ,in:

[0084] Location coordinates

[0085] Altitude information

[0086] Load information

[0087] Rope tension

[0088] Grab torque

[0089] Motor current

[0090] The observation probability density function is:

[0091] ;

[0092] 1.3 Hidden Markov Model (HMM)

[0093] The operation process of a bridge grab crane can be described using a five-element group. express:

[0094] State set ,

[0095] Observation set

[0096] State transition matrix ,in

[0097] Observation probability matrix ,in

[0098] Initial state probability vector

[0099] 1.4 State Transition Probability Model

[0100] Define the state transition probability matrix :

[0101] ;

[0102] Among them, according to the constraints of the work process:

[0103] (The ready state must be switched to the outbound state)

[0104] (The outbound process must be switched to the fetch-drop process.)

[0105] (The grab-drop state must be transitioned to the grab-close state)

[0106] (The grab-close state must be transitioned to the grab-lift state)

[0107] (The grab-promotion state must be transitioned to the return state)

[0108] (The return trip must be switched to the unloading-lowering state)

[0109] (The material feeding-lowering state must be switched to the material feeding-open state)

[0110] (The material feeding - open state must be switched to the material feeding - lifting state)

[0111] (The material feeding and lifting state must be switched to the ready state)

[0112] 1.5 Observation Probability Distribution Model

[0113] For each state The observation probability distribution adopts a multivariate Gaussian distribution:

[0114] ;

[0115] in:

[0116] :state The observed mean vector;

[0117] :state The observed covariance matrix;

[0118] : The dimension of the observation vector;

[0119] 2. Model Analysis and Algorithm

[0120] 2.1 Time Series Segmentation Algorithm

[0121] 2.1.1 Hidden Markov Model (HMM)-based segmentation algorithm

[0122] Input: Observation sequence ;

[0123] Output: State sequence ;

[0124] Algorithm steps:

[0125] 1) Initialize HMM parameters: ;

[0126] 2) Forward algorithm calculation:

[0127] ;

[0128] initialization: , ;

[0129] Recursion: ;

[0130] 3) Viterbi algorithm decoding:

[0131] ;

[0132] Path records: ;

[0133] 4) Optimal path backtracking:

[0134] ;

[0135] ;

[0136] 2.1.2 Refined Action Recognition Algorithm

[0137] Scraping process identification:

[0138] Python

[0139] edit

[0140] def detect_grabbing_process(X_segment):

[0141] Identify the three sub-stages of the crawling process

[0142] # 1. Decentralization Phase Testing

[0143] lowering_indices = detect_lowering_phase(X_segment)

[0144] # 2. Closure Phase Detection

[0145] closing_indices = detect_closing_phase(X_segment)

[0146] # 3. Enhancement Phase Testing

[0147] raising_indices = detect_raising_phase(X_segment)

[0148] return {

[0149] 'lowering': lowering_indices,

[0150] 'closing': closing_indices,

[0151] 'raising': raising_indices

[0152] }

[0153] def detect_lowering_phase(X):

[0154] Testing and decentralization phase

[0155] Conditions: Decreasing height, increasing load

[0156] h_diff = np.diff(X[:, height_idx])

[0157] w_diff = np.diff(X[:, weight_idx])

[0158] # Altitude Descent Conditions

[0159] height_condition = h_diff < -threshold_h

[0160] # Conditions for increased load

[0161] weight_condition = w_diff > threshold_w

[0162] return find_consecutive_true(height_condition)

[0163] def detect_closing_phase(X):

[0164] Detection closure phase

[0165] Conditions: High stability, sudden increase in grab torque

[0166] h_var = np.var(X[:, height_idx])

[0167] m_change = np.diff(X[:, moment_idx])

[0168] # High stability conditions

[0169] stable_condition = h_var < threshold_var

[0170] # Conditions for sudden increase in torque

[0171] moment_condition = m_change > threshold_m

[0172] return find_consecutive_true(stable_condition & moment_condition)

[0173] 2.2 Mathematical Model for Efficiency Evaluation

[0174] 2.2.1 Time Efficiency Model

[0175] Define time efficiency metrics:

[0176] Work cycle time:

[0177] ;

[0178] in:

[0179] Outbound travel time : Capture time, Return time Feeding time

[0180] Time breakdown by stage:

[0181] ;

[0182] Time utilization rate:

[0183] ;

[0184] 2.2.2 Space Efficiency Model

[0185] Path optimization rate:

[0186] ;

[0187] Displacement efficiency:

[0188] ;

[0189] 2.2.3 Energy Efficiency Model

[0190] Unit operating energy consumption:

[0191] ;

[0192] Energy efficiency ratio:

[0193] ;

[0194] Comprehensive efficiency model:

[0195] ;

[0196] in These are the weighting coefficients.

[0197] 3. Model Solving and Optimization Algorithms

[0198] 3.1 Parameter Estimation Algorithm

[0199] 3.1.1 Baum-Welch Algorithm

[0200] An expectation-maximization algorithm for estimating HMM parameters:

[0201] E-step (expected calculation):

[0202] ;

[0203] ;

[0204] M-step (parameter update):

[0205] ;

[0206] ;

[0207] ;

[0208] 3.2 Optimize the objective function

[0209] 3.2.1 Multi-objective optimization model

[0210] ;

[0211] in:

[0212] (Minimize job cycle)

[0213] (Minimize total energy consumption)

[0214] (Minimize operational volatility)

[0215] Constraints:

[0216] ;

[0217] ;

[0218] 3.2.2 Solving using a genetic algorithm

[0219] Chromosome coding:

[0220] ;

[0221] Fitness function:

[0222] ;

[0223] in .

[0224] 4. Model Validation and Simulation

[0225] 4.1 Simulation Experiment Design

[0226] Simulation parameter settings:

[0227] Simulation time: 30 days

[0228] Number of job cycles: 50,000

[0229] Data sampling frequency: 10Hz

[0230] State transition probability: based on historical data statistics

[0231] Simulation process:

[0232] 1. Initialize state → Generate observation sequence → State identification → Efficiency calculation → Result analysis.

[0233] 4.2 Model Performance Evaluation

[0234] 4.2.1 State recognition accuracy

[0235] As shown in the table below:

[0236]

[0237] 4.2.2 Efficiency Improvement Analysis

[0238] Improved time efficiency:

[0239] ;

[0240] Improved energy efficiency:

[0241] .

[0242] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any modifications or equivalent changes made based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A method for optimizing a bridge grab crane operation process based on time series data, characterized in that the method include: S1. Collect real-time time series data of the bridge grab crane during operation; S2. Based on the real-time time series data and hidden Markov model in step S1, construct the state space model of the work process. ,in For a set of states, For the observation set, Here is the state transition matrix. For the observation probability matrix, This is the initial state probability vector; S3. Based on real-time time series data, the forward algorithm and Viterbi algorithm are used to identify the status of the work process and segment the time series, and output the status sequence of the work stage. S4. Based on the identified state sequence, calculate the time efficiency index, spatial efficiency index and energy consumption efficiency index respectively, and construct a comprehensive efficiency evaluation model. S5. Estimate the parameters of the hidden Markov model based on the Baum-Welch algorithm, and combine the genetic algorithm to perform multi-objective optimization of the operation cycle, energy consumption and operation fluctuation, and output the optimized operation control parameters.

2. The method for optimizing the operation process of a bridge grab crane based on time series data according to claim 1, characterized in that, The real-time time series data in step S1 includes location coordinates, height information, load information, rope tension, grab torque, and motor current.

3. The method for optimizing the operation process of a bridge grab crane based on time series data according to claim 1, characterized in that, In step S2: State set The states are, in sequence, preparation state, outbound state, grab-and-release state, grab-and-close state, grab-and-lift state, return state, release-and-release state, release-and-open state, and release-and-lift state. Observation set , This includes location coordinates, altitude information, load information, rope tension, grab torque, and motor current; State transition matrix ,in ; Observation probability matrix ,in ; Initial state probability vector .

4. The method for optimizing the operation process of a bridge grab crane based on time series data according to claim 1, characterized in that, The comprehensive efficiency evaluation model in step S4 is expressed as follows: ； in, Indicates time efficiency. Indicates space efficiency. Indicates energy efficiency. , , Let represent the weighting coefficients, and + + =1.

5. The method for optimizing the operation process of a bridge grab crane based on time series data according to claim 1, characterized in that, The multi-objective optimization of the operation cycle, energy consumption, and operational fluctuations in step S5 is represented as follows: ； in: This indicates minimizing the job cycle. This represents minimizing total energy consumption. This indicates minimizing operational fluctuations.

6. The method for optimizing the operation process of a bridge grab crane based on time series data according to claim 5, characterized in that, The fitness function of the genetic algorithm in step S5 is as follows: ； in: .

Images