Abnormal lifting detection method, device and computer equipment of hoisting equipment

Abstract

The application relates to an abnormal lifting detection method and device of a lifting device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring current data and voltage data of the lifting device at multiple sampling time points; acquiring the value of a motor static parameter; determining the value of a first variable parameter at the multiple sampling time points according to the value of the motor static parameter, the current data and the voltage data; the motor static parameter is a parameter reflecting the internal properties of a motor in the lifting device; the first variable parameter is a parameter representing the lifting state of a heavy object; a first preset fitting function is acquired, which reflects the fitting relationship between the first variable parameter and time; and a first abnormal detection result of the lifting device is determined based on the first preset fitting function and according to the value of the first variable parameter at the multiple sampling time points. The method can improve the loading and unloading efficiency, simplify the detection process, reduce the investment cost and eliminate the environmental influence.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, apparatus, computer equipment, storage medium, and computer program product for detecting abnormal lifting of lifting equipment. Background Technology

[0002] Lifting equipment plays an increasingly important role in modern society, providing convenience for production and construction. However, with its widespread application across various industries, a serious problem has also emerged: during the lifting process, lifting equipment may lift vehicles carrying heavy loads along with the loads themselves, potentially leading to vehicle damage, equipment malfunction, and a series of other safety hazards. Therefore, it is essential to conduct inspections of the lifting process.

[0003] However, current inspections of lifting equipment during the lifting process often rely on the combined use of laser, infrared, and camera equipment. In addition to obvious drawbacks such as complex inspection processes, high investment costs, and susceptibility to environmental influences, these methods also require the equipment to be suspended for a period of time after being lifted, which cannot meet the needs of real-time inspection and severely restricts loading and unloading efficiency. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for detecting abnormal lifting of lifting equipment, which can improve loading and unloading efficiency, simplify the detection process, reduce investment costs, and eliminate environmental impact, in order to address the above-mentioned technical problems.

[0005] Firstly, this application provides a method for detecting abnormal lifting of lifting equipment. The method includes:

[0006] When a lifting device lifts a heavy object located on a loading object, acquire current and voltage data of the lifting device at multiple sampling time points;

[0007] The values ​​of the motor static parameters are obtained, and the values ​​of the first variable parameter at the multiple sampling time points are determined based on the values ​​of the motor static parameters, the current data, and the voltage data; the motor static parameters are parameters reflecting the internal properties of the motor in the lifting equipment; the first variable parameter is a parameter characterizing the lifting state of the heavy object.

[0008] Obtain a first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight;

[0009] Based on the first preset fitting function and according to the values ​​of the first variable parameter at the multiple sampling time points, the first anomaly detection result of the lifting equipment is determined.

[0010] In one embodiment, the motor static parameters include the number of power supply phases, motor rotor resistance, motor air gap power, and conversion coefficient; the first variable parameter includes lifting height and lifting weight; determining the value of the first variable parameter at the plurality of sampling time points based on the values ​​of the motor static parameters, the current data, and the voltage data includes:

[0011] The three-phase current and three-phase voltage are converted into effective current and effective voltage. Based on at least one of the current data or voltage data, the synchronous angular frequency of the motor is determined. Based on the effective current, effective voltage, stator resistance, and core loss resistance, the air gap power of the motor is determined. The ratio of the air gap power to the synchronous angular frequency of the motor is determined, and this ratio represents the motor torque. Based on the effective current, effective voltage, number of power supply phases, rotor resistance, stator resistance, and air gap power, the motor slip is determined. Based on the air gap power and slip, the motor output power is determined. Based on the synchronous angular frequency, number of poles, and slip, the motor speed is determined. Based on the conversion coefficient, the sampling time interval, and the sum of the motor speeds within the sampling time interval corresponding to the multiple sampling time points, the lifting height is determined. A second preset fitting function is obtained, and the motor output power and motor speed are input into the second preset fitting function to determine the lifting weight.

[0012] In one embodiment, the current data is the value of three current phases; the voltage data is the value of three voltage phases; determining the value of the motor synchronous angular frequency based on at least one of the current data or the voltage data includes:

[0013] Find multiple zero-crossing points among the values ​​of the three-phase current or the three-phase voltage; determine the time difference between two adjacent zero-crossing points based on the multiple zero-crossing points; take the reciprocal of the time difference to determine the value of the motor synchronous angular frequency.

[0014] In one embodiment, the first variable parameter includes at least one of lifting height and lifting weight; determining the first anomaly detection result of the lifting equipment based on the first preset fitting function and according to the values ​​of the first variable parameter at the plurality of sampling time points includes:

[0015] For each of the multiple sampling time points, the value of the first variable parameter at the target sampling time point is input into the first preset fitting function, and it is determined whether the first preset fitting function with the input value of the first variable parameter is valid; when the first preset fitting function with the input value of the first variable parameter is invalid, a first abnormality detection result characterizing the abnormal lifting of the lifting equipment is generated.

[0016] In one embodiment, when the first anomaly detection result indicates that the lifting equipment abnormally lifts the heavy object, the values ​​of the second variable parameter at the plurality of sampling time points are obtained; the second variable parameter is a parameter characterizing the internal operating state of the motor; a data grayscale image is generated based on the values ​​of the first variable parameter, the second variable parameter, the current data, and the voltage data at the plurality of sampling time points; the second anomaly detection result of the lifting equipment is determined by the abnormal lifting detection neural network and based on the data grayscale image.

[0017] In one embodiment, the second variable parameter includes at least one of the following: motor output power, motor torque, motor frequency, motor speed, and motor slip.

[0018] In one embodiment, generating a grayscale image based on the values ​​of a first variable parameter, a second variable parameter, the current data, and the voltage data at the plurality of sampling time points includes:

[0019] Obtain an initial grayscale image matrix; determine the first sorting relationship of the first variable parameter, the second variable parameter, current, and voltage in the columns of the initial grayscale image matrix based on the conversion relationship between the first variable parameter, the second variable parameter, current, and voltage; determine the second sorting relationship of the multiple sampling time points in the rows of the initial grayscale image matrix based on the temporal relationship of the multiple sampling time points; fill the initial grayscale image matrix with the current data, voltage data, values ​​of the first variable parameter, and values ​​of the second variable parameter at the multiple sampling time points according to the first sorting relationship and the second sorting relationship to obtain a target grayscale image matrix; generate a data grayscale image based on the target grayscale image matrix.

[0020] In one embodiment, determining the second anomaly detection result of the lifting equipment based on the grayscale image data via an abnormal lifting detection neural network includes:

[0021] The abnormal lifting detection neural network converts the grayscale image into a feature vector of a preset dimension; determines the polar radius of the feature vector in a preset polar coordinate system; and generates a second abnormality detection result characterizing the abnormal lifting of the lifting equipment when the polar radius is less than the preset polar radius threshold.

[0022] In one embodiment, determining the polar radius of the feature vector in a preset polar coordinate system includes:

[0023] Determine the vector components of the feature vector in multiple preset dimensions; square the vector components of the feature vector in each preset dimension to obtain the square value of each vector component; sum the square values ​​of each vector component to obtain the accumulated value; take the square root of the accumulated value to obtain the polar radius of the feature vector in the preset polar coordinate system.

[0024] In one embodiment, the training step of the abnormal lifting detection neural network includes:

[0025] Obtain a neural network to be trained; through the neural network to be trained, perform feature extraction on multiple unlabeled grayscale positive image samples and multiple labeled grayscale positive image samples to obtain a first sample feature vector of each unlabeled grayscale positive image sample in each preset dimension and a second sample feature vector of each labeled grayscale positive image sample in each preset dimension; adjust the neural network to be trained according to the first sample feature vector of each unlabeled grayscale positive image sample in each preset dimension and the second sample feature vector of each labeled grayscale positive image sample in each preset dimension to obtain an abnormal lifting detection neural network.

[0026] In one embodiment, adjusting the neural network to be trained based on the first sample feature vector of each unlabeled grayscale image positive sample in each preset dimension and the second sample feature vector of each labeled grayscale image positive sample in each preset dimension to obtain an anomaly detection neural network includes:

[0027] For each of multiple preset dimensions, the average value of multiple first sample feature vectors and multiple second sample feature vectors in the targeted dimension is determined; based on the average value of the multiple first sample feature vectors and multiple second sample feature vectors in the targeted dimension, the average vector of the preset number of dimensions is determined; the average vector component corresponding to each preset dimension is determined; the average vector component corresponding to each preset dimension is squared to obtain the sum of squares corresponding to each average vector component; the sum of squares corresponding to each average vector component is accumulated to obtain the accumulated sum of squares of the average vector components; the square root of the accumulated sum of squares of the average vector components is taken to obtain the origin coordinates corresponding to the average vector in the polar coordinate system; a loss function is constructed for the neural network to be trained, which is used to characterize the distance between the polar coordinates corresponding to the positive samples of the labeled grayscale image processed by the neural network to be trained and the origin coordinates, or the reciprocal of the distance between the polar coordinates corresponding to the negative samples of the labeled grayscale image and the origin coordinates; the neural network to be trained is adjusted according to the loss function to obtain the abnormal lifting detection neural network.

[0028] In one embodiment, adjusting the neural network to be trained according to the loss function to obtain an abnormal lifting detection neural network includes:

[0029] A target sample set is obtained, comprising multiple labeled positive grayscale image samples and multiple labeled negative grayscale image samples. At least one labeled positive or negative grayscale image sample is randomly selected from the target sample set as a target sample and input into the neural network to be trained. One of the selected target samples is processed. If the target sample is a labeled positive grayscale image sample, the neural network to be trained processes the labeled positive grayscale image sample to obtain a fourth feature vector. Based on the fourth feature vector, the polar radius of the labeled positive grayscale image sample in the polar coordinate system is determined, and the neural network to be trained is adjusted based on the polar radius. If the target sample is a labeled negative grayscale image sample, the neural network to be trained processes the labeled negative grayscale image sample to obtain a fifth feature vector. Based on the fifth feature vector, the reciprocal of the polar radius of the labeled negative grayscale image sample in the polar coordinate system is determined, and the neural network to be trained is adjusted based on the reciprocal of the polar radius.

[0030] In one embodiment, the step of generating the neural network to be trained includes:

[0031] Obtain unlabeled grayscale image samples and an initial neural network; extract features from the unlabeled grayscale image samples using the encoder of the initial neural network to obtain low-dimensional features; perform image reconstruction processing on the low-dimensional features using the decoder of the initial neural network to obtain a reconstructed image; adjust the parameters of the encoder and the decoder according to the difference between the reconstructed image and the unlabeled data samples to obtain the neural network to be trained.

[0032] Secondly, this application also provides an abnormal lifting detection device for lifting equipment. The device includes:

[0033] The first acquisition module is used to acquire current data and voltage data of the lifting equipment at multiple sampling time points when the lifting equipment lifts a heavy object located on the loading object;

[0034] The first determining module is used to acquire the values ​​of the motor static parameters, and determine the values ​​of the first variable parameter at the multiple sampling time points based on the values ​​of the motor static parameters, the current data, and the voltage data; the motor static parameters are parameters reflecting the internal properties of the motor in the lifting equipment; the first variable parameter is a parameter characterizing the lifting state of the heavy object;

[0035] The second acquisition module is used to acquire a first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight.

[0036] The second determining module is used to determine the first abnormality detection result of the lifting equipment based on the first preset fitting function and according to the value of the first variable parameter at the multiple sampling time points.

[0037] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0038] When a lifting device lifts a heavy object located on a loading object, acquire current and voltage data of the lifting device at multiple sampling time points;

[0039] The values ​​of the motor static parameters are obtained, and the values ​​of the first variable parameter at the multiple sampling time points are determined based on the values ​​of the motor static parameters, the current data, and the voltage data; the motor static parameters are parameters reflecting the internal properties of the motor in the lifting equipment; the first variable parameter is a parameter characterizing the lifting state of the heavy object.

[0040] Obtain a first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight;

[0041] Based on the first preset fitting function and according to the values ​​of the first variable parameter at the multiple sampling time points, the first anomaly detection result of the lifting equipment is determined.

[0042] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0043] When a lifting device lifts a heavy object located on a loading object, acquire current and voltage data of the lifting device at multiple sampling time points;

[0044] The values ​​of the motor static parameters are obtained, and the values ​​of the first variable parameter at the multiple sampling time points are determined based on the values ​​of the motor static parameters, the current data, and the voltage data; the motor static parameters are parameters reflecting the internal properties of the motor in the lifting equipment; the first variable parameter is a parameter characterizing the lifting state of the heavy object.

[0045] Obtain a first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight;

[0046] Based on the first preset fitting function and according to the values ​​of the first variable parameter at the multiple sampling time points, the first anomaly detection result of the lifting equipment is determined.

[0047] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0048] When a lifting device lifts a heavy object located on a loading object, acquire current and voltage data of the lifting device at multiple sampling time points;

[0049] The values ​​of the motor static parameters are obtained, and the values ​​of the first variable parameter at the multiple sampling time points are determined based on the values ​​of the motor static parameters, the current data, and the voltage data; the motor static parameters are parameters reflecting the internal properties of the motor in the lifting equipment; the first variable parameter is a parameter characterizing the lifting state of the heavy object.

[0050] Obtain a first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight;

[0051] Based on the first preset fitting function and according to the values ​​of the first variable parameter at the multiple sampling time points, the first anomaly detection result of the lifting equipment is determined.

[0052] The aforementioned abnormal lifting detection method, device, computer equipment, storage medium, and computer program product for lifting equipment, when the lifting equipment begins to lift a heavy object located on the loading object, first acquires current and voltage data of the lifting equipment at multiple sampling time points, and uses the acquired current and voltage data as a basis to detect the lifting status of the lifting equipment. This eliminates the need for auxiliary equipment such as lasers, infrared sensors, and cameras, simplifying the detection process, reducing unnecessary cost investment, and to some extent eliminating the influence of the environment on auxiliary equipment. Then, it acquires parameters of the internal properties of the motor in the lifting equipment, and based on the specific values ​​of these parameters, as well as the acquired current and voltage data, determines parameters related to the lifting status of the heavy object. Finally, the determined parameters related to the lifting status of the heavy object are input into a pre-set first preset fitting function to obtain the detection result of whether the lifting equipment is in an abnormal state during the lifting process. Since the acquisition and calculation processes of the aforementioned relevant data are real-time, they can meet the needs of real-time detection, greatly improving the loading and unloading efficiency of the lifting equipment. Attached Figure Description

[0053] Figure 1 This is a diagram illustrating the application environment of an abnormal lifting detection method for lifting equipment in one embodiment.

[0054] Figure 2 This is a flowchart illustrating an abnormal lifting detection method for a lifting device in one embodiment;

[0055] Figure 3 This is a schematic diagram of the physical model framework in one embodiment;

[0056] Figure 4 This is a schematic diagram illustrating the calculation and image processing of physical quantities in one embodiment.

[0057] Figure 5 This is a schematic diagram of semi-supervised neural network inference in one embodiment;

[0058] Figure 6 This is a flowchart illustrating an abnormal lifting detection method for lifting equipment in another embodiment;

[0059] Figure 7 This is a structural block diagram of an abnormal lifting detection device for a lifting device in one embodiment;

[0060] Figure 8 This is a structural block diagram of an abnormal lifting detection device for a lifting device in another embodiment;

[0061] Figure 9 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0063] The abnormal lifting detection method for lifting equipment provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or located in the cloud or on other network servers. Terminal 102 generates an abnormal lifting detection request for the lifting equipment and then sends this request to server 104 so that server 104 can determine the first abnormal detection result of the lifting equipment. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart vehicle devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0064] In one embodiment, such as Figure 2 As shown, a method for detecting abnormal lifting of lifting equipment is provided, which is applied to... Figure 1 Taking 104 as an example, the explanation includes the following steps:

[0065] Step 202: When the lifting equipment lifts the heavy object located on the loading object, acquire the current data and voltage data of the lifting equipment at multiple sampling time points.

[0066] This application does not specifically limit the lifting equipment; it can be set according to actual needs and can include cranes, elevators, hoists, etc. The loading object is used to load heavy objects, and there is no clear limitation on the specific type of loading object; it can include loading vehicles, loading containers, loading equipment, etc. Similarly, the types of heavy objects can be diverse, such as containers, cargo, etc.

[0067] Specifically, considering the needs of real-time performance and energy saving, the start time for abnormal lifting detection of the lifting equipment is set to the moment when the lifting equipment has just lifted the heavy object on the loading object. This moment is used as the start time for abnormal lifting detection, and the acquisition of current and voltage data of the lifting equipment at each sampling time point begins. It is easy to understand that the sampling time points can be set according to actual needs. For example, current and voltage data can be acquired continuously within a certain time period, or they can be acquired once at regular intervals. In other words, the sampling time points can be either discrete or continuous. This application does not limit the specific form of the current and voltage data; it can be three-phase current and three-phase voltage data, DC current and three-phase voltage data, or other forms of AC current and AC voltage data.

[0068] In one example, reference Figure 3 When the lifting equipment is detected to be running, the system begins to collect the necessary data via sensor 302. This includes obtaining the real-time three-phase current and three-phase voltage of the main drive motor of the lifting equipment. During this process, the algorithm filters out the instances where the wire rope is not taut during the lifting process, retaining only the data after the lifting device is subjected to the tension of the wire rope.

[0069] Step 204: Obtain the values ​​of the motor static parameters. Based on the values ​​of the motor static parameters, current data, and voltage data, determine the values ​​of the first variable parameter at multiple sampling time points. The motor static parameters are parameters that reflect the internal properties of the motor in the lifting equipment. The first variable parameter is a parameter that characterizes the lifting state of the heavy object.

[0070] Among them, the static parameters of the motor refer to the parameters that reflect the internal properties of the motor in the lifting equipment. The specific parameters included in the static parameters of the motor can be set according to actual needs, and may include, but are not limited to, the number of power supply phases, the motor rotor resistance, the motor air gap power, and the conversion factor. The first variable parameter refers to the parameter that characterizes the lifting state of the heavy object. The specific parameters included in the first variable parameter can be set according to actual needs, and may include, but are not limited to, the lifting weight and the lifting height.

[0071] Specifically, based on the actual needs of abnormal lifting detection, parameters reflecting the internal properties of the motor in the lifting equipment are acquired. Then, for multiple sampling time points, using the current data, voltage data, and parameters reflecting the internal properties of the motor in the lifting equipment at each sampling time point as the basic data, calculations are performed to obtain parameters characterizing the lifting state of the heavy object at that sampling time point.

[0072] In one example, reference Figure 3To extract more effective and intuitive information, it is necessary to model the lifting process of the heavy object. Motor model 304 is used to convert instantaneous current and voltage sampling values ​​into quantities that are easier to detect. When unloading a heavy object from the load, the lifting device primarily provides the pulling force for the object's ascent. Ignoring the influence of other mechanisms and devices such as frequency converters and reducers, a lifting motor model that provides power to the lifting device is considered. Using the lifting motor model, the motor speed, motor output power, electromagnetic torque, and other physical quantities can be calculated sequentially by inputting the current and voltage at specific sampling time points, and finally, the lifting height and lifting weight can be calculated.

[0073] Step 206: Obtain the first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight.

[0074] Specifically, a fitting function is a mathematical function used to approximate a set of data. The selection of a fitting function is usually based on the characteristics of the data and the expected fitting accuracy. A fitting relationship refers to using a fitting function to express and describe the relationship between actual data points. Through the fitting relationship, mathematical methods can be used to find the most suitable parameter values, enabling the fitting function to predict and approximate the given data as accurately as possible. The goal of the fitting relationship is to reduce the error between the data and the fitting function, thereby obtaining a good description and predictive ability of the data. In this application, the first preset fitting function is a function reflecting the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the load. It is easy to understand that in this first preset fitting function, both the first variable parameter and time can be considered as variables in the first preset fitting function. It is easy to understand that the content of the first preset fitting function is not specifically limited and can be set according to actual needs.

[0075] In one example, when the load is not separated from the object being lifted, or when it is lifted along with the object, the lifting process is also related to the suspension system of the object being lifted; therefore, modeling of the suspension system is also considered. A complete suspension system includes guiding mechanisms, elastic elements, trapezoidal structures, damping elements, stabilizing devices, and limiting blocks. In abnormal lifting detection of hoisting equipment, the longitudinal displacement change of the object being lifted is often the primary concern; therefore, the suspension model only considers the parts that have the greatest impact on longitudinal displacement: elastic elements, damping elements, and wheels. The suspension system of the object being lifted is simplified to a spring-damped system.

[0076] Step 208: Based on the first preset fitting function and according to the values ​​of the first variable parameter at multiple sampling time points, determine the first anomaly detection result of the lifting equipment.

[0077] In this application, the first abnormal detection result refers to the lifting equipment lifting the load object along with the load during the lifting process.

[0078] Specifically, refer to Figure 3 For multiple sampling time points, firstly, based on the current data, voltage data, and parameters reflecting the internal properties of the motor in the lifting equipment at each sampling time point, calculations are performed to obtain the parameter characterizing the lifting state of the load at that sampling time point, i.e., the value of the first variable parameter at each sampling time point. Then, the values ​​of the first variable parameters corresponding to each sampling time point, obtained through calculations, are substituted into a first preset fitting function using the suspension model 306. This first preset fitting function reflects the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the load. Based on the first preset fitting function after substituting specific values, an abnormal detection result is determined: whether the lifting equipment lifted the loaded object along with the load during the lifting process.

[0079] In one step, the three-phase current and three-phase voltage are converted into effective current and effective voltage. Based on at least one of the current or voltage data, the synchronous angular frequency of the motor is determined. Based on the effective current, effective voltage, stator resistance, and core loss resistance, the air gap power of the motor is determined. The ratio of the air gap power to the synchronous angular frequency is determined, and this ratio represents the motor torque. Based on the effective current, effective voltage, number of power supply phases, rotor resistance, stator resistance, and air gap power, the motor slip is determined. Based on the air gap power and slip, the motor output power is determined. Based on the synchronous angular frequency, number of poles, and slip, the motor speed is determined. Based on the conversion coefficient, the sampling time interval, and the sum of the motor speeds within the sampling time interval corresponding to multiple sampling time points, the lifting height is determined. A second preset fitting function is obtained, and the motor output power and motor speed are input into the second preset fitting function to determine the lifting weight.

[0080] In this context, the number of power phases refers to the number of power sources in a power system, usually expressed numerically. Common power phase numbers include single-phase, two-phase, and three-phase. The motor rotor resistance refers to the resistance contained within the motor rotor. Rotor resistance is determined by the resistive properties of the rotor conductor material and affects the motor's current and torque characteristics. The motor air gap power refers to the power generated by the motor through the air gap between the rotor and stator during operation. When the motor rotor rotates, due to the tiny gap between the rotor and stator, magnetic field exchange occurs, converting electrical energy into air gap magnetic field energy; this energy is the air gap power. The conversion factor in this application refers to the conversion factor between the motor speed and the wire rope linear speed. The first variable parameter includes lifting height and lifting weight, where lifting height refers to the height to which the load is lifted by the lifting equipment, and lifting weight refers to the weight lifted by the lifting equipment during the lifting process.

[0081] Specifically, the synchronous angular frequency of a motor refers to the angular frequency at which the motor rotates under ideal conditions. It can also be understood as the rotational speed of the motor rotor when the motor's trigger signals are perfectly synchronized. The synchronous angular frequency can be obtained by statistically analyzing multiple zero-crossing points in the three-phase current values ​​or multiple zero-crossing points in the three-phase voltage values.

[0082] In one example, the conversion between motor frequency and motor angular frequency is: Motor angular frequency (ωs) = 2π × Motor frequency (f). Here, the motor angular frequency is expressed in radians per second, and the motor frequency is expressed in Hertz (Hz). In other words, multiplying the motor frequency by 2π gives the motor angular frequency. Conversely, to convert the motor angular frequency to the motor frequency, simply divide the motor angular frequency by 2π. For example, if the motor angular frequency is 60 radians per second, then dividing 60 by 2π gives approximately 9.549 Hertz.

[0083] Motor torque refers to the torque generated by a motor during rotation, used to describe the magnitude and direction of the motor's rotational force. The value of motor torque can be obtained by determining the ratio of the motor's air gap power to its synchronous angular frequency. The value of the motor's air gap power can be calculated from the effective voltage, effective current, stator resistance, and core loss resistance through certain operational relationships. Stator resistance refers to the total resistance of the motor's stator windings, the resistance encountered when the conductors pass through the stator windings. Core loss resistance, also known as iron loss resistance or magnetic loss resistance, refers to the resistance encountered when the magnetic energy generated by the core material in a motor or transformer is converted into heat energy.

[0084] In one example, it can be calculated using the formula: T = Pgap / Ws. Where Pgap is the air gap power of the motor, T is the motor torque, and Ws is the synchronous angular frequency of the motor.

[0085] Pgap can be expressed by the formula: PI 2 ×Rs-e 2 Rc is calculated. Where P is active power, I is effective current, Rs is stator resistance, and Rc is core loss resistance. e is the input voltage after removing the stator resistance voltage drop, which can be calculated using the formula: e = VI × Rs. Where V is effective voltage, I is effective current, and Rs is stator resistance.

[0086] Motor slip refers to the difference between the actual rotor speed and the synchronous speed of a motor during operation; it can also be called slip rate. The value of motor slip can be calculated from the motor's air gap power, rotor resistance, number of power supply phases, effective current, effective voltage, and stator resistance through certain operational relationships. Among these, the motor rotor resistance, also known as rotor winding resistance, refers to the resistance of the windings wound on the motor rotor.

[0087] In one example, the motor slip can be calculated using the formula: s = (Pgap × Rr) / (q × e²). Where Rr is the motor rotor resistance, q is the number of power supply phases, and s is the motor slip.

[0088] Motor output power refers to the mechanical power generated by a motor per unit time, usually expressed in watts (W). It represents the efficiency with which the motor converts electrical energy into mechanical energy. The value of motor output power can be calculated from the motor power value and the motor slip value through a certain calculation relationship.

[0089] In one example, the motor output power can be calculated using the formula: Pout = Pgap × (1-s), where s is the motor slip and Pout is the motor output power.

[0090] Motor speed refers to the angular velocity of a motor's rotation, usually expressed in revolutions per minute (rpm) or revolutions per second (r / s). It is a physical quantity that measures the speed of a motor's rotation. The value of motor speed can be calculated by using certain arithmetic relationships between the motor's synchronous angular frequency, the number of poles, and the slip. The number of poles refers to the number of pole pairs on the motor rotor, which can also be understood as the number of magnetic poles. It is an important indicator in motor design and operation.

[0091] In one example, the motor speed can be calculated using the formula: n = 60Ws × (1-s) / 2π × p. Where p is the number of motor poles, s is the motor slip, Ws is the motor synchronous angular frequency, and n is the motor speed.

[0092] Lifting height refers to the height to which a heavy object is lifted by hoisting equipment. The value of lifting height can be calculated by combining the conversion coefficient, the interval between sampling time points, and the sum of the motor speeds during that interval, through a specific calculation relationship.

[0093] In one example, the lifting height can be calculated using the formula: h = c / Ts × ∑n(s). Here, c is the conversion coefficient between the motor speed and the wire rope linear velocity, Ts is the sampling time interval, and h is the lifting height.

[0094] The lifting weight refers to the weight lifted by the lifting equipment during the lifting process. The lifting weight value can be obtained by inputting the motor output power value and the motor speed value into a preset second preset fitting function. After the second preset fitting function calculates, the lifting weight value is obtained.

[0095] In one example, the weight to be lifted can be calculated using the formula: m = f(Pout, n). Here, n is the motor speed, Pout is the motor output power, f is the second preset fitting function, and m is the weight to be lifted.

[0096] In one embodiment, multiple zero-crossing points are found among the values ​​of the three-phase current or the three-phase voltage; the time difference between two adjacent zero-crossing points is determined based on the multiple zero-crossing points; and the reciprocal of the time difference is taken to determine the value of the motor synchronous angular frequency.

[0097] Zero-crossing refers to the instant when a current or voltage changes from a positive to a negative value or vice versa. In sinusoidal alternating current (AC), both current and voltage periodically cross zero-crossing points. Zero-crossing is a characteristic of AC, reflecting the direction and change of current or voltage. In motors, the occurrence of zero-crossing points is related to the motor's rotation direction and phase. When a motor is operating, current and voltage periodically cross zero-crossing points, allowing the rotor to rotate continuously. For induction motors and AC permanent magnet motors, zero-crossing is a necessary condition for normal operation.

[0098] Specifically, first, determine the specific values ​​of the three-phase current or three-phase voltage and find multiple zero-crossing points. Devices such as sensors or oscilloscopes can be used to monitor the waveforms of the current or voltage in real time and record the times of multiple zero-crossing points. Once multiple zero-crossing points are found, the period of the current or voltage can be determined by calculating the time difference between two adjacent zero-crossing points. The time difference can be calculated from the time difference between adjacent zero-crossing points. Then, the reciprocal of these time differences can be taken to determine the synchronous angular frequency of the motor. The synchronous angular frequency refers to the angular velocity of the motor's rotation, which is related to the motor's operating frequency and number of poles. The value of the synchronous angular frequency can be calculated by dividing the reciprocal of the time difference by 2π (pi).

[0099] By measuring multiple zero-crossing points and calculating the time difference between adjacent zero-crossing points, a relatively accurate value of the motor's synchronous angular frequency can be obtained. This helps in understanding the motor's operating status and speed. Furthermore, by monitoring the waveform of current or voltage in real time and recording the times of multiple zero-crossing points, the motor's synchronous angular frequency can be acquired in real time. This helps in timely understanding changes in the motor's operating state.

[0100] In one embodiment, for each of the multiple sampling time points, the value of the first variable parameter at the target sampling time point is input to a first preset fitting function, and it is determined whether the first preset fitting function with the input value of the first variable parameter is valid; when the first preset fitting function with the input value of the first variable parameter is invalid, a first abnormality detection result characterizing the abnormal lifting of the lifting equipment is generated.

[0101] This application does not limit the specific content of the first preset fitting function, which can be set according to actual needs. It is easy to understand that the first preset fitting function is mainly for a specific sampling time point, used to determine whether the lifting equipment at that sampling time point is in an abnormal lifting state.

[0102] Specifically, at each sampling time point, the value of the first variable parameter is input into the first preset fitting function, and it is determined whether the preset function is valid. If the first preset fitting function corresponding to the value of the first variable parameter is invalid, it indicates that an anomaly exists. At this time, a first anomaly detection result characterizing the abnormal lifting of the lifting equipment is generated. This result may include a detailed description of the anomaly or other indications to help operators determine whether there is a problem with the lifting equipment.

[0103] By utilizing pre-defined fitting functions and algorithms, an automated anomaly detection process can be achieved, reducing human intervention and the subjectivity of judgment. This improves the accuracy and reliability of detection. Furthermore, by analyzing data from multiple sampling time points simultaneously, anomalies at multiple time points can be detected concurrently. This batch detection method improves detection efficiency and reduces the waste of human resources.

[0104] In one embodiment, when the first anomaly detection result indicates that the lifting equipment is abnormally lifting a heavy object, the values ​​of the second variable parameter at multiple sampling time points are obtained; the second variable parameter is a parameter that characterizes the internal operating state of the motor; a data grayscale image is generated based on the values ​​of the first variable parameter, the second variable parameter, current data, and voltage data at multiple sampling time points; the second anomaly detection result of the lifting equipment is determined by the abnormal lifting detection neural network and based on the data grayscale image.

[0105] The second variable parameter is a parameter characterizing the internal operating state of the motor. It is readily understood that this application does not limit the specific content of the second variable parameter, and it can be set according to actual needs. The grayscale image can be understood as a two-dimensional image composed of multiple parameters, obtained by stacking the values ​​of the first and second variable parameters on the vertical axis. In this two-dimensional image, the horizontal axis represents time, and the vertical axis represents different first and second variable parameters.

[0106] In this study, considering that the initial anomaly detection result obtained through the first preset fitting function is not accurate enough, a second detection using an anomaly detection neural network is required. This second detection complements the first detection during operation, and the final determination of an anomaly is made only if both detections identify the lifting equipment as being in an abnormal position.

[0107] The specific process includes: firstly, acquiring the values ​​of a first and second variable parameter at multiple sampling time points; generating a grayscale image by combining current and voltage data; and then using an abnormal lifting detection neural network to determine the second anomaly detection result of the lifting equipment. The first variable parameter is primarily used to determine abnormal lifting conditions of the lifting equipment. When an anomaly occurs, corresponding anomaly information is generated, such as abnormal acceleration, velocity, or displacement data. The second variable parameter is mainly used to characterize the internal operating state of the motor. Changes in the internal operating state of the motor may lead to abnormal operation of the lifting equipment; therefore, by monitoring the value of the second variable parameter, it is possible to further determine whether there is a problem with the equipment.

[0108] Secondly, by comprehensively considering the first variable parameter, the second variable parameter, current data, and voltage data, a grayscale image is generated. The grayscale image is a visual image formed based on the numerical distribution of different parameters, which can intuitively reflect the operating status and abnormal conditions of the equipment.

[0109] Finally, the grayscale images of the data are input into the abnormal lifting detection neural network for further analysis. The neural network can learn and recognize patterns of different working and abnormal states, thereby determining whether there is a secondary abnormality in the lifting equipment.

[0110] First, a first-stage detection is performed using a pre-defined fitting function, followed by a second detection using an anomaly detection neural network. This approach fully leverages the interpretability of the pre-defined fitting function and the high accuracy of the anomaly detection neural network, thereby reducing false alarms, preventing missed detections, ensuring staff safety, and minimizing economic losses.

[0111] In one embodiment, the second variable parameter includes at least one of the following: motor output power, motor torque, motor frequency, motor speed, and motor slip.

[0112] Motor output power refers to the mechanical power generated by the motor per unit time, usually expressed in watts (W). Motor torque refers to the torque generated by the motor during rotation, used to describe the magnitude and direction of the motor's rotational force. Motor frequency refers to the number of rotational cycles the motor completes per second. In AC motors, the voltage supplied by the power source is achieved by periodically changing its direction, and the motor changes its speed and direction according to the voltage changes. Motor frequency is usually expressed in Hertz (Hz), representing the number of cycles per second. Motor speed refers to the angular velocity of the motor's rotation, usually expressed in revolutions per minute (rpm) or revolutions per second (r / s). Motor slip is the difference between the actual rotor speed and the synchronous speed during motor operation; it can also be called slip rate.

[0113] In one embodiment, an initial grayscale image matrix is ​​obtained; based on the conversion relationship between the first variable parameter, the second variable parameter, current, and voltage, a first sorting relationship is determined in the columns of the initial grayscale image matrix; based on the temporal relationship of multiple sampling time points, a second sorting relationship is determined in the rows of the initial grayscale image matrix; based on the first and second sorting relationships, the current data, voltage data, values ​​of the first variable parameter, and values ​​of the second variable parameter at multiple sampling time points are filled into the initial grayscale image matrix to obtain a target grayscale image matrix; and a data grayscale image is generated based on the target grayscale image matrix.

[0114] Specifically, refer to Figure 4 First, collect physical quantity and time data: You need to collect a set of physical quantities and their corresponding time data. This could be experimental data, sensor data, etc. Next, create an image matrix: Create a two-dimensional grayscale image matrix based on the size of your dataset. The number of rows in the matrix represents the time dimension, and the number of columns represents the number of physical quantities. Then, standardize the physical quantities: For each physical quantity, standardize its value to a commonly used range, such as 0 to 255. Ensure that different physical quantities are within the same numerical range. Next, fill the matrix according to correlations: Based on the correlations between physical quantities, fill the corresponding rows of adjacent physical quantities in the image matrix with the corresponding grayscale values. You can use color mapping tools to map values ​​to grayscale values. Finally, fill the matrix according to temporal relationships: Based on temporal continuity, fill the corresponding columns of physical quantity values ​​at adjacent time points in the image matrix with the corresponding grayscale values. Finally, plot the image: Use an image processing library to plot the generated image matrix as a grayscale image.

[0115] Traditional anomaly detection algorithms often process one-dimensional data. According to the scheme in this application, an input sample would have 960 dimensions (60 * 16 = 960). This massive dimensionality leads to the curse of dimensionality, not only drastically increasing computational cost but also causing extremely sparse data distribution in space, resulting in a significant decrease in accuracy whether processed directly or through dimensionality reduction. However, due to the local connectivity and parameter sharing characteristics of convolution, convolutional neural networks can process even high-resolution images quickly without falling into the curse of dimensionality. Therefore, we also choose convolution to handle massive amounts of data. We concatenate the voltage data, current data, first variable parameter, and second variable parameter obtained during acquisition and calculation, and convert them into an image that can be processed by a convolutional neural network.

[0116] In one embodiment, an abnormal lifting detection neural network is used to convert a grayscale image into a feature vector of a preset dimension; the polar radius of the feature vector in a preset polar coordinate system is determined; and if the polar radius is less than a preset polar radius threshold, a second abnormality detection result representing the abnormal lifting of the lifting equipment is generated.

[0117] Specifically, refer to Figure 5 First, the grayscale image data is converted into a feature vector of a preset dimension. The anomaly lift detection neural network can extract image features through operations such as convolution and pooling, mapping them to the feature vector. Next, the polar radius of the feature vector in a preset polar coordinate system is determined. The polar radius represents the distance between the feature vector and the origin, and this distance can be used to quantify the degree of anomaly of the feature vector. The larger the polar radius, the greater the difference between the feature vector and the normal state. Then, the polar radius is compared with a preset polar radius threshold. If the polar radius is smaller than the preset polar radius threshold, it means that the grayscale image data corresponding to the feature vector represents an abnormal lift. The anomaly lift detection neural network can be trained to acquire the ability to identify lift anomalies and generate a second anomaly detection result accordingly.

[0118] Converting feature vectors to extreme radii simplifies high-dimensional feature vectors into a one-dimensional extreme radius value, further reducing feature complexity and dimensionality. Compared to the original high-dimensional feature vectors, extreme radius calculation and comparison are more efficient, accelerating the anomaly detection process, improving detection efficiency, and thus reducing computational and storage costs. Furthermore, after converting feature vectors to extreme radii, preset extreme radius thresholds can be set according to specific application needs to determine anomaly scenarios. This simple numerical comparison makes threshold setting more intuitive and easier.

[0119] In one embodiment, the vector components of the feature vector in multiple preset dimensions are determined; the vector components of the feature vector in each preset dimension are squared to obtain the square value corresponding to each vector component; the square values ​​corresponding to each vector component are accumulated to obtain the accumulated value; the square root of the accumulated value is taken to obtain the polar radius of the feature vector in the preset polar coordinate system.

[0120] Specifically, refer to Figure 5 Feature vectors are typically high-dimensional. First, we need to determine the components of the feature vector in each preset dimension. Next, we square each vector component in each preset dimension to obtain the squared value of each vector component in the preset dimension. Then, we sum the squared values ​​to obtain a cumulative value. Finally, we square-root the cumulative value to obtain the final extreme radius value.

[0121] The feature vector is converted into its polar radius in polar coordinates. This conversion simplifies feature representation and extracts key information from the data while reducing dimensionality and computational complexity. The polar radius reflects the weight and influence of the feature vector's components across different preset dimensions, which helps in identifying abnormal lifting situations.

[0122] In one embodiment, a neural network to be trained is obtained; through the neural network to be trained, features are extracted from multiple unlabeled grayscale positive samples and multiple labeled grayscale positive samples to obtain a first sample feature vector of each unlabeled grayscale positive sample in each preset dimension and a second sample feature vector of each labeled grayscale positive sample in each preset dimension; based on the first sample feature vector of each unlabeled grayscale positive sample in each preset dimension and the second sample feature vector of each labeled grayscale positive sample in each preset dimension, the neural network to be trained is adjusted to obtain an abnormal lifting detection neural network.

[0123] Specifically, refer to Figure 5First, a neural network model is needed, typically a predefined model, such as a common deep learning model like a Convolutional Neural Network (CNN). The neural network to be trained is used to extract features from multiple unlabeled and labeled positive grayscale images. Feature extraction is the process of converting an image into a set of feature vectors, each representing a feature of the image in a preset dimension. Then, for each unlabeled and labeled positive grayscale image, features are extracted using the neural network, resulting in a feature vector for each sample in each preset dimension. This feature vector is a numerical vector, with each element representing a feature component of the sample in the corresponding preset dimension. Finally, using the feature vectors obtained in the previous step, the neural network to be trained is adjusted based on the differences between the unlabeled and labeled positive samples. After multiple rounds of parameter tuning and training, an anomaly detection neural network is finally obtained. This network is trained in multiple preset dimensions based on the differences in feature vectors between unlabeled and labeled samples, with the aim of effectively distinguishing between normal and anomalous suspension situations.

[0124] In one embodiment, for each of a plurality of preset dimensions, the average value of a plurality of first sample feature vectors and a plurality of second sample feature vectors in the targeted dimension is determined; based on the average value of the plurality of first sample feature vectors and a plurality of second sample feature vectors in the targeted dimension, the average vector of a preset number of dimensions is determined; the average vector component corresponding to each preset dimension is determined; the average vector component corresponding to each preset dimension is squared to obtain the sum of squares corresponding to each average vector component; the sum of squares corresponding to each average vector component is accumulated to obtain the accumulated sum of squares of the average vector components; the square root of the accumulated sum of squares of the average vector components is taken to obtain the origin coordinates of the average vector in the polar coordinate system; a loss function is constructed for the neural network to be trained, which is used to characterize the distance between the polar coordinates of the labeled grayscale positive sample and the origin coordinates in the polar coordinate system after processing by the neural network to be trained, or the reciprocal of the distance between the polar coordinates of the labeled grayscale negative sample and the origin coordinates in the polar coordinate system; the neural network to be trained is adjusted according to the loss function to obtain the abnormal lifting detection neural network.

[0125] Specifically, refer to Figure 5The average output of positive grayscale images is calculated based on unlabeled and labeled positive grayscale images in the training set. This average output is a 256-dimensional vector, which represents the center of all positive grayscale images. We will use this vector as a proxy for the center of all positive grayscale images. During training, this center is fixed, and the normality of each input sample is measured by the Euclidean distance between its output (encoded) and this center. During training, we use the reciprocal of the distances from labeled positive grayscale images to the center and the distances from labeled negative grayscale images to the center as the loss function. The optimization objective is to minimize the loss function; therefore, during optimization, the network will move normal points closer to the center and outliers away from the center, thus completing the classification.

[0126] By using the reciprocals of the distances from positive samples to the center of the labeled grayscale image and the distances from negative samples to the center of the labeled grayscale image as loss functions, the network's optimization objective is to minimize the loss function. In this way, during training, the network will bring normal samples closer to the center and move abnormal samples further away from the center, thereby improving classification performance.

[0127] In one embodiment, a target sample set is obtained, which includes multiple labeled positive grayscale image samples and multiple labeled negative grayscale image samples. At least one labeled positive or negative grayscale image sample is arbitrarily selected from the target sample set as a target sample and input into the neural network to be trained. One of the selected target samples is processed. If the target sample is a labeled positive grayscale image sample, the neural network to be trained processes the labeled positive grayscale image sample to obtain a fourth feature vector. Based on the fourth feature vector, the polar radius of the labeled positive grayscale image sample in the polar coordinate system is determined, and the neural network to be trained is adjusted accordingly. If the target sample is a labeled negative grayscale image sample, the neural network to be trained processes the labeled negative grayscale image sample to obtain a fifth feature vector. Based on the fifth feature vector, the reciprocal of the polar radius of the labeled negative grayscale image sample in the polar coordinate system is determined, and the neural network to be trained is adjusted accordingly.

[0128] Specifically, refer to Figure 5First, a sample is selected from the target sample set and input into the neural network to be trained. This sample can be either a positive or negative sample. If the input is a positive sample (i.e., a target object in a grayscale image), the neural network will calculate a result called the fourth feature vector. Based on this feature vector, the polar radius of the positive sample in polar coordinates can be determined. The polar radius reflects the position of the positive sample in feature space. Then, the parameters in the neural network to be trained are adjusted according to this polar radius. If a negative sample (i.e., a non-target object in a grayscale image) is selected, the neural network to be trained will also calculate a result called the fifth feature vector. Then, based on this feature vector, the polar radius of the negative sample in polar coordinates can be determined, and its reciprocal is taken. Taking the reciprocal is to distinguish the influence of the negative sample from that of the positive sample during training. Then, the parameters in the neural network to be trained are adjusted according to the reciprocal of this polar radius.

[0129] By using labeled positive and negative samples as training data, this scheme employs supervised learning, which effectively guides the neural network to learn to distinguish between target and non-target objects. Furthermore, the neural network to be trained is adjusted based on the polar radius of the positive samples and the inverse of the polar radius of the negative samples. This flexible adjustment strategy helps the model to gradually optimize during training, improving its final classification performance on unknown data.

[0130] In one embodiment, unlabeled grayscale image samples and an initial neural network are acquired; the encoder of the initial neural network extracts features from the unlabeled grayscale image samples to obtain low-dimensional features; the decoder of the initial neural network performs image reconstruction processing on the low-dimensional features to obtain a reconstructed image; based on the difference between the reconstructed image and the unlabeled data samples, the parameters of the encoder and decoder are adjusted to obtain the neural network to be trained.

[0131] Specifically, the encoder transforms the input unlabeled grayscale image samples (raw data) and compresses them into a low-dimensional representation, thereby reducing the use of computing resources and storage space.

[0132] The decoder is responsible for receiving the encoder's output and restoring it to reconstructed data. Similar to the encoder, the decoder consists of a series of linear layers, deconvolutional layers, etc., with the output of the last layer having the same dimension as the unlabeled grayscale image sample.

[0133] During training, we define the Mean Squared Error (MSE) loss to measure the difference between the reconstructed and original data. During training, the encoder attempts to minimize the MSE loss, i.e., minimize the difference between the original and reconstructed data, thus enabling the encoder and decoder to learn a good low-dimensional representation and decoding rule. Finally, the output of the intermediate layer in the encoder is the feature representation of the original data. In summary, a set of compression and decoding rules with low information loss is learned, thus fully preserving the information in the original input.

[0134] By using MSE loss as a metric, the encoder and decoder can learn an effective data compression and decompression rule, thereby achieving efficient data processing and storage while preserving key information in the original data.

[0135] In one embodiment, such as Figure 6 As shown, Figure 6 A flowchart illustrating an abnormal lifting detection method for lifting equipment in another embodiment is provided, including:

[0136] Step 602: When the lifting equipment lifts the heavy object located on the loading object, acquire the current data and voltage data of the lifting equipment at multiple sampling time points;

[0137] Step 604: Locate multiple zero-crossing points among the values ​​of the three-phase current or three-phase voltage; determine the time difference between two adjacent zero-crossing points based on the multiple zero-crossing points; take the reciprocal of the time difference to determine the value of the motor synchronous angular frequency; determine the value of the motor air gap power based on the values ​​of effective current, effective voltage, stator resistance, and core loss resistance; determine the ratio of the motor air gap power to the motor synchronous angular frequency, which is the value of the motor torque; determine the value of the motor slip based on the values ​​of effective current, effective voltage, number of power supply phases, motor rotor resistance, stator resistance, and motor air gap power; determine the value of the motor output power based on the values ​​of the motor air gap power and motor slip; determine the value of the motor speed based on the values ​​of the motor synchronous angular frequency, number of motor poles, and motor slip; determine the value of the lifting height based on the value of the conversion coefficient, the interval of the sampling time points, and the sum of the motor speeds within the sampling time periods corresponding to multiple sampling time points; obtain the second preset fitting function, input the values ​​of the motor output power and motor speed into the second preset fitting function, and determine the value of the lifting weight;

[0138] Step 606: Obtain the first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight.

[0139] Step 608: For each of the multiple sampling time points, input the value of the first variable parameter at the corresponding sampling time point into the first preset fitting function, and determine whether the first preset fitting function with the input value of the first variable parameter is valid; when the first preset fitting function with the input value of the first variable parameter is invalid, generate a first abnormality detection result characterizing the abnormal lifting of the lifting equipment.

[0140] Step 610: When the first anomaly detection result indicates that the lifting equipment is abnormally lifting a heavy object, obtain the values ​​of the second variable parameter at multiple sampling time points; the second variable parameter is a parameter characterizing the internal operating state of the motor, and the second variable parameter includes at least one of the following: motor output power, motor torque, motor frequency, motor speed, and motor slip; obtain the initial grayscale image matrix; determine the first sorting relationship of the first variable parameter, the second variable parameter, current, and voltage in the columns of the initial grayscale image matrix according to the conversion relationship between the first variable parameter, the second variable parameter, current, and voltage; determine the second sorting relationship of the multiple sampling time points in the rows of the initial grayscale image matrix according to the temporal relationship of the multiple sampling time points; and sort the multiple sampling time points according to the first and second sorting relationships. The current data, voltage data, and values ​​of the first and second variable parameters at each sampling time point are filled into the initial grayscale image matrix to obtain the target grayscale image matrix. A data grayscale image is generated based on the target grayscale image matrix. The data grayscale image is converted into a feature vector of a preset dimension through an abnormal lifting detection neural network. The vector components of the feature vector in multiple preset dimensions are determined. The vector components of the feature vector in each preset dimension are squared to obtain the square value corresponding to each vector component. The square values ​​corresponding to each vector component are accumulated to obtain the accumulated value. The square root of the accumulated value is taken to obtain the polar radius of the feature vector in a preset polar coordinate system. If the polar radius is less than a preset polar radius threshold, a second abnormality detection result representing the abnormal lifting of the lifting equipment is generated.

[0141] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0142] Based on the same inventive concept, this application also provides an abnormal lifting detection device for lifting equipment, which implements the abnormal lifting detection method for lifting equipment described above. The solution provided by this device is similar to the implementation described in the above method. Therefore, the specific limitations of one or more embodiments of the abnormal lifting detection device for lifting equipment provided below can be found in the limitations of the abnormal lifting detection method for lifting equipment described above, and will not be repeated here.

[0143] In one embodiment, such as Figure 7 As shown, an abnormal lifting detection device 700 for lifting equipment is provided, comprising: a first acquisition module 702, a first determination module 704, a second acquisition module 706, and a second determination module 708, wherein:

[0144] The first acquisition module 702 is used to acquire current data and voltage data of the lifting equipment at multiple sampling time points when the lifting equipment lifts a heavy object located on the loading object.

[0145] The first determining module 704 is used to obtain the values ​​of the motor static parameters, and determine the values ​​of the first variable parameter at multiple sampling time points based on the values ​​of the motor static parameters, current data, and voltage data; the motor static parameters are parameters that reflect the internal properties of the motor in the lifting equipment; the first variable parameter is a parameter that characterizes the lifting state of the heavy object;

[0146] The second acquisition module 706 is used to acquire a first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight.

[0147] The second determining module 708 is used to determine the first abnormality detection result of the lifting equipment based on the first preset fitting function and the values ​​of the first variable parameter at multiple sampling time points.

[0148] In one embodiment, the first determining module 704 is used to convert three-phase current and three-phase voltage into effective current and effective voltage; determine the value of the motor synchronous angular frequency based on at least one of the current data or voltage data; determine the value of the motor air gap power based on the values ​​of effective current, effective voltage, stator resistance, and core loss resistance; determine the ratio of the motor air gap power to the motor synchronous angular frequency, where the ratio is the value of the motor torque; determine the value of the motor slip based on the values ​​of effective current, effective voltage, number of power supply phases, motor rotor resistance, stator resistance, and motor air gap power; determine the value of the motor output power based on the values ​​of the motor air gap power and motor slip; determine the value of the motor speed based on the values ​​of the motor synchronous angular frequency, number of motor poles, and motor slip; determine the value of the lifting height based on the value of the conversion coefficient, the interval of sampling time points, and the sum of the motor speeds within the sampling time period corresponding to multiple sampling time points; and obtain a second preset fitting function, inputting the values ​​of the motor output power and motor speed into the second preset fitting function to determine the value of the lifting weight.

[0149] In one embodiment, the first determining module 704 is used to find multiple zero-crossing points among the values ​​of the three-phase current or the three-phase voltage; determine the time difference between two adjacent zero-crossing points based on the multiple zero-crossing points; and determine the value of the motor synchronous angular frequency by taking the reciprocal of the time difference.

[0150] In one embodiment, the second determining module 708 is used to input the value of the first variable parameter at each of the multiple sampling time points into a first preset fitting function, and determine whether the first preset fitting function with the input value of the first variable parameter is valid; when the first preset fitting function with the input value of the first variable parameter is not valid, a first abnormality detection result characterizing the abnormal lifting of the lifting equipment is generated.

[0151] In one embodiment, the abnormal lifting detection device for the lifting equipment further includes a neural network detection module 710, used to acquire the values ​​of a second variable parameter at multiple sampling time points when the first abnormal detection result indicates that the lifting equipment is abnormally lifting a heavy object; the second variable parameter is a parameter that characterizes the internal operating state of the motor; a data grayscale image is generated based on the values ​​of the first variable parameter, the second variable parameter, current data, and voltage data at multiple sampling time points; and the second abnormal detection result of the lifting equipment is determined by the abnormal lifting detection neural network and based on the data grayscale image.

[0152] In one embodiment, the neural network detection module 710 is used to acquire an initial grayscale image matrix; determine a first sorting relationship of the first variable parameter, the second variable parameter, the current, and the voltage in the columns of the initial grayscale image matrix according to the conversion relationship between the first variable parameter, the second variable parameter, and the current and voltage; determine a second sorting relationship of the multiple sampling time points in the rows of the initial grayscale image matrix according to the temporal relationship of the multiple sampling time points; fill the initial grayscale image matrix with the current data, voltage data, the values ​​of the first variable parameter, and the values ​​of the second variable parameter at the multiple sampling time points according to the first sorting relationship and the second sorting relationship to obtain a target grayscale image matrix; and generate a data grayscale image based on the target grayscale image matrix.

[0153] In one embodiment, the neural network detection module 710 is used to convert the grayscale image of the data into a feature vector of a preset dimension through an abnormal lifting detection neural network; determine the polar radius of the feature vector in a preset polar coordinate system; and generate a second abnormality detection result characterizing the abnormal lifting of the lifting equipment when the polar radius is less than a preset polar radius threshold.

[0154] In one embodiment, the neural network detection module 710 is used to determine the vector components of the feature vector in multiple preset dimensions; to square the vector components of the feature vector in each preset dimension to obtain the square value corresponding to each vector component; to accumulate the square values ​​corresponding to each vector component to obtain the accumulated value; and to take the square root of the accumulated value to obtain the polar radius of the feature vector in the preset polar coordinate system.

[0155] In one embodiment, the abnormal lifting detection device for the lifting equipment further includes a training module 712 for acquiring a neural network to be trained; through the neural network to be trained, feature extraction is performed on multiple unlabeled grayscale positive image samples and multiple labeled grayscale positive image samples to obtain a first sample feature vector of each unlabeled grayscale positive image sample in each preset dimension and a second sample feature vector of each labeled grayscale positive image sample in each preset dimension; based on the first sample feature vector of each unlabeled grayscale positive image sample in each preset dimension and the second sample feature vector of each labeled grayscale positive image sample in each preset dimension, the neural network to be trained is adjusted to obtain the abnormal lifting detection neural network.

[0156] In one embodiment, the training module 712 is configured to: determine the average value of multiple first sample feature vectors and multiple second sample feature vectors in each of multiple preset dimensions; determine the average vector of a preset number of dimensions based on the average value of the multiple first sample feature vectors and multiple second sample feature vectors in the targeted dimension; determine the average vector component corresponding to each preset dimension; square the average vector component corresponding to each preset dimension to obtain the sum of squares corresponding to each average vector component; accumulate the sum of squares corresponding to each average vector component to obtain the accumulated sum of squares of the average vector components; take the square root of the accumulated sum of squares of the average vector components to obtain the origin coordinates of the average vector in the polar coordinate system; construct a loss function for the neural network to be trained, the loss function being used to characterize the distance between the polar coordinates of the labeled grayscale positive sample and the origin coordinates in the polar coordinate system after processing by the neural network to be trained, or the reciprocal of the distance between the polar coordinates of the labeled grayscale negative sample and the origin coordinates in the polar coordinate system; and adjust the neural network to be trained according to the loss function to obtain an anomaly detection neural network.

[0157] In one embodiment, the training module 712 is used to acquire a target sample set, which includes multiple labeled grayscale positive image samples and multiple labeled grayscale negative image samples; at least one labeled grayscale positive image sample or labeled grayscale negative image sample is arbitrarily selected from the target sample set as a target sample and input into the neural network to be trained; one of the selected target samples is processed; if the target sample is a labeled grayscale positive image sample, the labeled grayscale positive image sample is processed by the neural network to be trained to obtain a fourth feature vector; based on the fourth feature vector, the polar radius of the polar coordinates corresponding to the labeled grayscale positive image sample in the polar coordinate system is determined, and the neural network to be trained is adjusted according to the polar radius; if the target sample is a labeled grayscale negative image sample, the labeled grayscale negative image sample is processed by the neural network to be trained to obtain a fifth feature vector; based on the fifth feature vector, the reciprocal of the polar radius of the polar coordinates corresponding to the labeled grayscale negative image sample in the polar coordinate system is determined, and the neural network to be trained is adjusted according to the reciprocal of the polar radius.

[0158] In one embodiment, the training module 712 is used to acquire unlabeled grayscale image samples and an initial neural network; to extract features from the unlabeled grayscale image samples using the encoder of the initial neural network to obtain low-dimensional features; to perform image reconstruction processing on the low-dimensional features using the decoder of the initial neural network to obtain a reconstructed image; and to adjust the parameters of the encoder and decoder according to the difference between the reconstructed image and the unlabeled data samples to obtain the neural network to be trained.

[0159] In another embodiment, such as Figure 8 As shown, Figure 8 The structural block diagram of an abnormal lifting detection device for hoisting equipment in another embodiment includes: a first acquisition module 702, a first determination module 704, a second acquisition module 706, and a second determination module 708. The abnormal lifting detection device 700 further includes a neural network detection module 710 and a training module 712.

[0160] Each module in the abnormal lifting detection device of the aforementioned lifting equipment can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0161] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores data related to abnormal lifting detection of the lifting equipment. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for abnormal lifting detection of lifting equipment.

[0162] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0163] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0164] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0165] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0166] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0167] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0168] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0169] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for detecting abnormal lifting of lifting equipment, characterized in that, The method includes: When a lifting device lifts a heavy object located on a loading object, acquire current and voltage data of the lifting device at multiple sampling time points; The values ​​of the motor static parameters are obtained, and the values ​​of the first variable parameter at the multiple sampling time points are determined based on the values ​​of the motor static parameters, the current data, and the voltage data; the motor static parameters are parameters reflecting the internal properties of the motor in the lifting equipment; the first variable parameter is a parameter characterizing the lifting state of the heavy object. Obtain a first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight; Based on the first preset fitting function and according to the values ​​of the first variable parameter at the multiple sampling time points, the first anomaly detection result of the lifting equipment is determined.

2. The method according to claim 1, characterized in that, The static parameters of the motor include the number of power phases, the rotor resistance, the air gap power, and the conversion coefficient; the first variable parameter includes the lifting height and the lifting weight. The step of determining the value of the first variable parameter at the plurality of sampling time points based on the values ​​of the motor static parameters, the current data, and the voltage data includes: The three-phase current and the three-phase voltage are converted into effective current and effective voltage, and the value of the motor synchronous angular frequency is determined based on at least one of the current data or the voltage data. Based on the effective current, the effective voltage, the stator resistance, and the core loss resistance, the value of the motor air gap power is determined, and the ratio of the motor air gap power to the motor synchronous angular frequency is determined, where the ratio is the value of the motor torque. The value of motor slip is determined based on the effective current, the effective voltage, the number of power supply phases, the value of motor rotor resistance, the value of stator resistance, and the value of motor air gap power. The value of the motor output power is determined based on the value of the motor air gap power and the value of the motor slip. The motor speed is determined based on the values ​​of the motor synchronous angular frequency, the number of motor poles, and the motor slip. The lifting height is determined based on the value of the conversion coefficient, the interval of the sampling time points, and the sum of the motor speeds within the sampling time period corresponding to the plurality of sampling time points; Obtain a second preset fitting function, input the value of the motor output power and the value of the motor speed into the second preset fitting function, and determine the value of the weight to be increased.

3. The method according to claim 2, characterized in that, The current data consists of three current values; the voltage data consists of three voltage values; determining the motor synchronous angular frequency based on at least one of the current data or the voltage data includes: Find multiple zero-crossing points among the values ​​of the three-phase current or the three-phase voltage; The time difference between two adjacent zero crossings is determined based on the plurality of zero crossings; The reciprocal of the time difference is used to determine the value of the motor's synchronous angular frequency.

4. The method according to claim 1, characterized in that, The first variable parameter includes at least one of lifting height and lifting weight; The step of determining the first anomaly detection result of the lifting equipment based on the first preset fitting function and according to the values ​​of the first variable parameter at the multiple sampling time points includes: For each of the plurality of sampling time points, the value of the first variable parameter at the target sampling time point is input into the first preset fitting function, and it is determined whether the first preset fitting function with the input value of the first variable parameter is valid; When the first preset fitting function with the first variable parameter value is not valid, a first anomaly detection result characterizing the abnormal lifting of the lifting equipment is generated.

5. The method according to claim 1, characterized in that, The method further includes: When the first anomaly detection result indicates that the lifting equipment abnormally lifts the heavy object, the value of the second variable parameter at the multiple sampling time points is obtained; the second variable parameter is a parameter characterizing the internal operating state of the motor. A grayscale image is generated based on the values ​​of the first variable parameter, the second variable parameter, the current data, and the voltage data at the multiple sampling time points; The second abnormality detection result of the lifting equipment is determined by using an abnormal lifting detection neural network and based on the grayscale image of the data.

6. The method according to claim 5, characterized in that, The second variable parameter includes at least one of the following: motor output power, motor torque, motor frequency, motor speed, and motor slip.

7. The method according to claim 5, characterized in that, The step of generating a grayscale image based on the values ​​of the first variable parameter, the second variable parameter, the current data, and the voltage data at the multiple sampling time points includes: Obtain the initial grayscale image matrix; Based on the conversion relationship between the first variable parameter, the second variable parameter, current, and voltage, determine the first sorting relationship of the first variable parameter, the second variable parameter, current, and voltage in the columns of the initial grayscale image matrix; Based on the temporal order of the multiple sampling time points, a second sorting relationship of the multiple sampling time points in the rows of the initial grayscale image matrix is ​​determined; Based on the first sorting relationship and the second sorting relationship, the current data, voltage data, the value of the first variable parameter and the value of the second variable parameter at the multiple sampling time points are filled into the initial grayscale image matrix to obtain the target grayscale image matrix. Generate a data grayscale image based on the target grayscale image matrix.

8. The method according to claim 5, characterized in that, The step of determining the second anomaly detection result of the lifting equipment based on the grayscale image data, using an abnormal lifting detection neural network, includes: The abnormal lifting detection neural network converts the grayscale image into a feature vector of a preset dimension. Determine the polar radius of the feature vector in the preset polar coordinate system; When the extreme diameter is less than the preset extreme diameter threshold, a second abnormality detection result is generated to characterize the abnormal lifting of the lifting equipment.

9. The method according to claim 8, characterized in that, Determining the polar radius of the feature vector in the preset polar coordinate system includes: Determine the vector components of the feature vector in multiple preset dimensions; The vector components of the feature vector in each preset dimension are squared to obtain the squared value of each vector component. The summation of the squared values ​​of each vector component is obtained by summing the summation values. The square root of the accumulated value is taken to obtain the polar radius of the feature vector in the preset polar coordinate system.

10. The method according to claim 5, characterized in that, The training steps of the abnormal lifting detection neural network include: Obtain the neural network to be trained; Through the neural network to be trained, features are extracted from multiple unlabeled grayscale positive samples and multiple labeled grayscale positive samples to obtain the first sample feature vector of each unlabeled grayscale positive sample in each preset dimension and the second sample feature vector of each labeled grayscale positive sample in each preset dimension. The neural network to be trained is adjusted based on the first sample feature vector of each unlabeled grayscale positive sample in each preset dimension and the second sample feature vector of each labeled grayscale positive sample in each preset dimension to obtain an abnormal lifting detection neural network.

11. The method according to claim 10, characterized in that, The step of adjusting the neural network to be trained based on the first sample feature vector of each unlabeled grayscale positive sample in each preset dimension and the second sample feature vector of each labeled grayscale positive sample in each preset dimension to obtain an abnormal lifting detection neural network includes: For each of the multiple preset dimensions, determine the average value of multiple first sample feature vectors and multiple second sample feature vectors in the targeted dimension; The average vector of a preset number of dimensions is determined based on the average of multiple first sample feature vectors and multiple second sample feature vectors in the targeted dimension. Determine the average vector components corresponding to each preset dimension of the average vector; Squaring the average vector component corresponding to each preset dimension yields the sum of squares for each average vector component. The sum of the squares of each average vector component is accumulated to obtain the sum of the squares of the average vector components. The square root of the sum of the squares of the average vector components is taken to obtain the coordinates of the origin of the average vector in the polar coordinate system. Construct a loss function for the neural network to be trained. The loss function is used to characterize the distance between the polar coordinates of the labeled grayscale positive sample in the polar coordinate system and the coordinates of the origin, or the reciprocal of the distance between the polar coordinates of the labeled grayscale negative sample in the polar coordinate system and the coordinates of the origin. The neural network to be trained is adjusted according to the loss function to obtain an abnormal lifting detection neural network.

12. The method according to claim 11, characterized in that, The step of adjusting the neural network to be trained according to the loss function to obtain the abnormal lifting detection neural network includes: Obtain a target sample set, which includes multiple labeled grayscale image positive samples and multiple labeled grayscale image negative samples; At least one labeled positive grayscale image sample or labeled negative grayscale image sample is randomly selected from the target sample set as the target sample and input into the neural network to be trained. One of the selected target samples is processed. If the target sample is a labeled grayscale positive sample, the labeled grayscale positive sample is processed by the neural network to be trained to obtain a fourth feature vector. Based on the fourth feature vector, the polar radius of the polar coordinates corresponding to the labeled grayscale positive sample in the polar coordinate system is determined. Based on the polar radius, the neural network to be trained is adjusted. If the target sample is a labeled grayscale image negative sample, the labeled grayscale image negative sample is processed by the neural network to be trained to obtain a fifth feature vector; based on the fifth feature vector, the reciprocal of the polar radius of the polar coordinates corresponding to the labeled grayscale image negative sample in the polar coordinate system is determined, and the neural network to be trained is adjusted based on the reciprocal of the polar radius.

13. The method according to claim 10, characterized in that, The generation steps of the neural network to be trained include: Obtain unlabeled grayscale image samples and an initial neural network; The encoder of the initial neural network extracts features from the unlabeled grayscale image samples to obtain low-dimensional features. The low-dimensional features are reconstructed using the decoder of the initial neural network to obtain a reconstructed image. Based on the difference between the reconstructed image and the unlabeled data sample, the parameters of the encoder and the decoder are adjusted to obtain the neural network to be trained.

14. An abnormal lifting detection device for lifting equipment, characterized in that, The device includes: The first acquisition module is used to acquire current data and voltage data of the lifting equipment at multiple sampling time points when the lifting equipment lifts a heavy object located on the loading object; The first determining module is used to acquire the values ​​of the motor static parameters, and determine the values ​​of the first variable parameter at the multiple sampling time points based on the values ​​of the motor static parameters, the current data, and the voltage data; the motor static parameters are parameters reflecting the internal properties of the motor in the lifting equipment; the first variable parameter is a parameter characterizing the lifting state of the heavy object; The second acquisition module is used to acquire a first preset fitting function; the first preset fitting function is used to reflect the fitting relationship between the first variable parameter and time when the loaded object is lifted off the ground along with the weight. The second determining module is used to determine the first abnormality detection result of the lifting equipment based on the first preset fitting function and according to the value of the first variable parameter at the multiple sampling time points.

15. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 13.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 13.

17. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 13.

Images