Hoist, hoist system, and state estimation device
The hoisting machine system uses learned state estimation formulas to estimate states like load mass and temperature without physical models, addressing operational challenges and enhancing safety for unskilled workers.
Patent Information
- Application Number
- JP2021062041
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-03-31
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2041-03-31
AI Technical Summary
Existing hoisting machines face challenges in accurately estimating states such as load mass, motor temperature, and load swing without constructing complex physical models, which are influenced by various factors like power transmission system deterioration and motor characteristics, making it difficult for unskilled workers to operate safely.
A hoisting machine system that includes a motor and a state estimation unit using learned state estimation formulas to estimate states based on motor operating information, eliminating the need for a physical model, and allowing for simultaneous calculation of multiple interrelated states.
Enables accurate estimation of hoisting machine states without complex physical models, reducing calculation load and costs, and enhancing safety by providing real-time operational insights for unskilled operators.
Smart Images

Figure 0007593865000001 
Figure 0007593865000002 
Figure 0007593865000003
Abstract
Description
[Technical field]
[0001] The present invention relates generally to a hoist. [Background technology]
[0002] In recent years, the number of inexperienced workers (unskilled workers) is increasing due to the aging of skilled hoist operators and labor shortages caused by the increase in the number of hoists installed. Therefore, there is a demand for hoists that can realize safe and highly efficient transportation even by unskilled workers. When operating a hoist, there are various needs, such as notifying the operator of various operating conditions including the mass of the load and supporting the transportation operation, preventing overload, predicting lifespan, diagnosing deterioration, suppressing load swing, and detecting ground clearance when the load leaves the ground.
[0003] In particular, when it comes to lifting the load from the ground, hoisting a point that is not aligned with the center of gravity of the load can cause dangerous situations, such as the load swinging wildly at the moment of lifting or the sling wire breaking, and this cannot be ignored in terms of ensuring the safety of the worker. For this reason, technology has been proposed that allows even unskilled workers to lift the load safely. In this regard, a technology has been disclosed that ensures safety by acquiring the current value and rotation speed of the motor that constitutes the hoisting machine and detecting the moment immediately before lifting the load using the results of comparing the two (see Patent Document 1). [Prior art documents] [Patent documents]
[0004] [Patent Document 1] JP 2012-66893 A Summary of the Invention [Problem to be solved by the invention]
[0005] According to the method described in Patent Document 1, ground-cutting can be detected using output information from the motor. However, since the motor current value is affected by multiple elements constituting the hoist, such as deterioration of the power transmission system of bearings and gears, and changes in motor temperature, in order to improve the accuracy of ground-cutting detection, it is necessary to precisely create a physical model by taking into account conditional branching. Furthermore, even when estimating various states related to the hoist, such as estimating the mass of the suspended load and estimating the temperature of the motor, in addition to ground-cutting, it is necessary to construct a physical model and set parameters for estimating the items to be subject to each state estimation (referred to as "estimated items"). Therefore, in state estimation using a physical model, both improving the estimation accuracy and increasing the number of estimated items have in common the fact that the factors of the physical items to be modeled increase, making it difficult to set parameters.
[0006] The present invention has been made in consideration of the above points, and aims to propose a hoist etc. that is capable of estimating the state related to the hoist without constructing a physical model. [Means for solving the problem]
[0007] In order to solve such problems, the present invention provides a hoisting machine that has a motor and controls the motor to hoist and lower a load, and that includes an operation information acquisition unit that acquires operation information of the motor, and a state estimation unit that estimates a state related to the hoisting machine based on the operation information acquired by the operation information acquisition unit and a state estimation equation that has been learned to be able to estimate a state related to the hoisting machine.
[0008] In the above configuration, the state of the hoisting machine can be estimated by using the motor operation information and the state estimation equation, without constructing a physical model. Effect of the Invention
[0009] According to the present invention, the state of the hoist can be estimated. [Brief description of the drawings]
[0010] [Figure 1] FIG. 1 is a diagram showing an example of a configuration of a hoist system according to a first embodiment. [Diagram 2] FIG. 2 is a diagram showing an example of a configuration of a hoist according to the first embodiment. [Diagram 3] FIG. 2 is a diagram showing an example of a configuration of a hoist according to the first embodiment. [Figure 4] FIG. 2 is a diagram showing an example of a configuration of a hoist according to the first embodiment. [Diagram 5] FIG. 2 is a diagram illustrating an example of a functional configuration of the hoist system according to the first embodiment. [Figure 6] FIG. 4 is a diagram showing an example of a state estimation equation according to the first embodiment. [Figure 7] FIG. 2 is a diagram showing an example of a flowchart according to the first embodiment. [Figure 8] FIG. 2 is a diagram showing an example of a flowchart according to the first embodiment. [Figure 9] FIG. 11 is a diagram showing an example of the configuration of a hoisting machine system according to a second embodiment. [Figure 10] FIG. 11 is a diagram showing an example of a flowchart according to the second embodiment. [Figure 11] FIG. 11 is a diagram showing an example of a flowchart according to the second embodiment. [Figure 12] FIG. 11 is a diagram showing an example of a flowchart according to the second embodiment. [Figure 13] FIG. 11 is a diagram showing an example of the configuration of a hoisting machine system according to a third embodiment. [Figure 14] FIG. 11 is a diagram illustrating an example of a flowchart according to the third embodiment. [Figure 15] FIG. 13 is a diagram showing an example of the configuration of a hoisting machine system according to a fourth embodiment. [Figure 16] FIG. 13 is a diagram showing an example of the configuration of a hoisting machine system according to a fifth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0011] (I) First embodiment Hereinafter, a mode for carrying out the present invention (hereinafter, referred to as an "embodiment") will be described in detail. However, the present invention is not limited to the embodiment.
[0012] In this embodiment, a hoist that includes a motor and controls the motor to hoist and lower a load is described as an example. For example, the hoist includes an operation information acquisition unit that acquires operation information of the motor, and a state estimation unit that estimates a state related to the hoist from the operation information acquired by the operation information acquisition unit. The state related to the hoist is the state of the motor that constitutes the hoist, the state of the hoist to which the load is attached, the state of the load, and the like. For example, the state estimation unit estimates at least one state related to the hoist based on a state estimation formula that is learned using learning data in which operation information and information indicating a state related to the hoist are associated with each other.
[0013] According to the above configuration, by using a state estimation formula based on appropriate learning, it becomes possible to predict various states related to the hoist from the operation information of the motor without constructing a physical model. Furthermore, according to the above configuration, it is possible to estimate the state related to the hoist in a complex system in which multiple factors affect each other. Furthermore, according to the above configuration, it is possible to simultaneously calculate estimated values of estimation items having a parallel relationship, such as having a common physical background, and / or estimated values of estimation items having a serial relationship, such as a dependency relationship.
[0014] Next, an embodiment of the present invention will be described based on the drawings. The following description and drawings are examples for explaining the present invention, and are omitted and simplified as appropriate for clarity of explanation. The present invention can be implemented in various other forms. Unless otherwise specified, each component may be singular or plural.
[0015] In the following description, the same elements in the drawings are given the same numbers, and the description is omitted as appropriate. When describing elements of the same type without distinction, the common part (part excluding the branch number) of the reference sign including the branch number is used, and when describing elements of the same type while distinguishing between them, the reference sign including the branch number may be used. For example, in this specification, when describing hoists without any particular distinction, they may be described as "hoist 101," and when describing individual hoists while distinguishing between them, they may be described as "hoist 101-1," "hoist 101-2," and so on.
[0016] In addition, the designations "first," "second," "third," and the like in this specification are given to identify components, and do not necessarily limit the number or order. Furthermore, numbers for identifying components are used in each context, and a number used in one context does not necessarily indicate the same configuration in another context. Furthermore, a component identified by a certain number is not prevented from also having the function of a component identified by another number.
[0017] The first embodiment will be described with reference to Figures 1 to 8. In Figure 1, 100 indicates a hoist system according to this embodiment as a whole.
[0018] 1 is a diagram showing an example of the configuration of a hoisting machine system 100. The hoisting machine system 100 includes a crane 110 having a hoisting machine 101 mounted thereon, and an operation terminal 120.
[0019] The crane 110 includes a runway 111 provided along both walls of a building (not shown), a girder 112 that moves on the runway 111, and a trolley 113 that moves along the girder 112. A conveying motor that drives the girder 112 and the trolley 113 is connected to the girder 112 and the trolley 113. A hoist 101 is provided on the trolley 113, and a rope suspended from the hoist 101 is slinged around a suspended load 130.
[0020] The term "rope" mentioned here is used as a general term for the suspending equipment used to suspend the load 130. In other words, "rope" includes not only so-called ropes, but also chains, belts, wires, cables, strings, cords, etc. Also, below, the machine related to the transportation of the load 130 may be referred to as a "product." The product includes at least the hoist 101. The product may be a crane 110, but depending on the product, it does not include the runway 111, the girder 112, the trolley 113, etc.
[0021] 2 to 4 are diagrams showing an example of the configuration of the hoist 101. The hoist 101 is configured to include a hoist drum 210 that winds up a rope, a motor 220 that is a power source that rotates the hoist drum 210, and a bearing 230 for smoothly transmitting power from the motor 220 to the hoist drum 210. In this embodiment, the motor 220 is assumed to be a three-phase electric motor, but since the effects of this embodiment do not depend on the type of motor 220, various motors 220 such as a DC motor can be applied. Moreover, the power from the motor 220 may be transmitted via a gear, a belt, or the like.
[0022] The transport motors installed on the girder 112 and the trolley 113, respectively, and the motor 220 installed on the hoisting machine 101 rotate in response to operation commands from the operation terminal 120. As a result, the hoisting machine 101 performs a hoisting operation and a lowering operation, and the girder 112 and the trolley 113 move in a specified direction, so that a series of transport operations are performed in which the load 130 is hoisted, transported to a specified position, and lowered. Note that the load 130 is an object to be transported by the crane 110, and is not a component of the crane 110.
[0023] Here, a description will be given of a representative state related to the hoist 101 in which state estimation is performed. At the time of lifting from the ground, the hoist system 100 estimates the mass of the suspended load 130 (referred to as "suspended load mass"), the probability that the rope is taut or slack (referred to as "tension state"), and the temperature of the motor 220 (referred to as "motor temperature"), and estimates the amount of load sway of the suspended load 130 (referred to as "load sway amount") immediately after lifting from the ground.
[0024] In addition, since the suspended load mass is strongly related to the load applied to the motor 220, the hoist system 100 can also estimate the state (force, acceleration, etc.) expressed by the relational expression of the second law of motion equation in addition to or instead of the suspended load mass. In other words, the hoist system 100 can also estimate the acceleration of the suspended load 130, the tension applied to the rope, the rotational load (load torque) of the motor 220 including the power transmission system, etc.
[0025] The reason why the present embodiment assumes the time of ground removal is that, at ground removal, the motor 220 of the hoisting machine 101 is mainly driven, and therefore, the state estimation is less likely to be influenced by other factors. However, the state estimation is not limited to the time of ground removal, and may be continuously performed during transportation and lowering. Ground removal refers to the state from when the load 130 shown in FIG. 2 is placed on the ground 240 to when the load 130 shown in FIG. 3 is separated from the ground 240.
[0026] Here, a case where a deviation occurs in the positional relationship between the load 130 and the trolley 113 when the load is removed from the ground will be described with reference to Fig. 4. When the load is removed from the ground, it is desirable for the trolley 113 to be positioned directly above the center of gravity 410 of the load 130, but as shown in Fig. 4, there are cases where the trolley 113 is positioned at a position offset from the center of gravity 410 of the load 130. In this case, the amount of load swing after the load is removed from the ground may become large.
[0027] FIG. 5 is a diagram showing an example of the configuration of a calculation processing system for control signals in the hoist system 100 (the functional configuration of the hoist system 100).
[0028] The operation terminal 120 includes a status display unit 521 and an operation input unit 522. The operation terminal 120 includes hardware such as a calculation unit 523 such as a CPU, a main storage unit 524 such as a semiconductor memory, an auxiliary storage unit 525 such as a hard disk, a communication unit 526, and an input / output unit 527 combining a display and a touch panel. The well-known operations of each piece of hardware will be omitted as appropriate in the following description. The operation terminal 120 may be a teaching pendant, or may be a terminal such as a smartphone, a tablet terminal, a notebook computer, or a desktop computer. The operation terminal 120 may be provided as a dedicated terminal, or the functions of the operation terminal 120 may be provided as an application and used on each operator's own terminal.
[0029] The functions of the operation terminal 120 (such as the status display unit 521 and the operation input unit 522) may be realized, for example, by the arithmetic unit 523 reading a program stored in the auxiliary storage unit 525 into the main storage unit 524 and executing it (software), or may be realized by hardware such as a dedicated circuit, or may be realized by a combination of software and hardware. Note that one function of the operation terminal 120 may be divided into multiple functions, or multiple functions may be integrated into one function. Also, some of the functions of the operation terminal 120 may be provided as a separate function, or may be included in another function. Also, some of the functions of the operation terminal 120 may be realized by another computer capable of communicating with the operation terminal 120.
[0030] The status display unit 521 displays, for example, an estimated value of the suspended load mass, an estimated value of the motor temperature, and an estimated value of the load swing amount as numerical values on the input / output device 527. In addition, the status display unit 521 displays, for example, an estimated value of the tension state as either "tension" or "relaxation" on the input / output device 527. However, the display by the status display unit 521 may be omitted as necessary.
[0031] The operation input unit 522 is an interface for inputting operation commands to the hoist 101 by the operator operating an input / output device 527 such as a button, a lever, a handle, a joystick, a keyboard, or a mouse. The operation input unit 522 may be an interface (hardware) combining a display and a touch panel, or an interface (software) realized on a device such as a smartphone. The operation input unit 522 may input a travel distance, coordinates of a destination, a location name, and the like, in response to an operation of the input / output device 527 by the operator. The operation input unit 522 may be linked to other systems such as an operation management system and a production management system that manage the hoist 101 and the like installed in a factory, and may input operation commands based on information from the other systems. Furthermore, the operation input unit 522 may be provided with an automatic operation function that automatically inputs pre-programmed operation commands.
[0032] In this embodiment, the status display unit 521 and the operation input unit 522 are implemented in the operation terminal 120, but may be implemented in the hoist 101. For example, the status display unit 521 may be provided on a board mounted on the hoist 101.
[0033] Next, the hoisting machine 101 will be described. The hoisting machine 101 includes a motor 220, a motor control unit 511, an operation information acquisition unit 512, and a state estimation unit 513. The hoisting machine 101 also includes hardware such as a calculation unit 514 such as a CPU, a main storage unit 515 such as a semiconductor memory, an auxiliary storage unit 516 such as a hard disk, and a communication unit 517.
[0034] The functions of the hoist 101 (motor control unit 511, operation information acquisition unit 512, state estimation unit 513, etc.) may be realized, for example, by the arithmetic device 514 reading a program stored in the auxiliary storage device 516 into the main storage device 515 and executing it (software), or may be realized by hardware such as a dedicated circuit, or may be realized by a combination of software and hardware. Note that one function of the hoist 101 may be divided into multiple functions, or multiple functions may be integrated into one function. Also, some of the functions of the hoist 101 may be provided as a separate function, or may be included in another function. Also, some of the functions of the hoist 101 may be realized by another computer capable of communicating with the hoist 101.
[0035] In this embodiment, data transmission by wireless communication is assumed, but the effect is the same even if communication is performed using a wired cable. For example, an operation command inputted to the operation terminal 120 is transmitted to the hoist 101 via the communication device 526 and the communication device 517.
[0036] The motor control unit 511 receives an operation command inputted via the operation input unit 522 of the operation terminal 120, generates a motor control command for moving the motor 220 according to the operation command, and outputs the motor control command to the motor 220. The motor control unit 511 also transmits a part or all of the motor control command outputted to the motor 220 to the driving information acquisition unit 512. However, the transmission of the motor control command from the motor control unit 511 to the driving information acquisition unit 512 may be performed via the motor 220.
[0037] The operation information acquisition unit 512 acquires operation information to be transmitted to the state estimation unit 513. The operation information is input / output information related to the operation of the motor 220 used for state estimation in the state estimation unit 513, and is information indicating a part or all of the motor control command. For example, the operation information is information such as a voltage value applied to the motor 220, a command frequency to a control unit (inverter) of the motor 220, a motor rotation speed calculated from a sensor value of an encoder provided for the motor 220, a drive current value input to the motor 220, an excitation current value and / or a torque current value in vector control of the motor 220, an input torque value to the motor 220, a command speed input to the motor 220, and a slip value of the rotation speed of the motor 220 relative to the command speed input to the motor 220.
[0038] The state estimation unit 513 estimates a state from the driving information acquired by the driving information acquisition unit 512 using a state estimation equation stored in the auxiliary storage device 516 .
[0039] In this embodiment, in the pre-shipment process of the hoist 101, data consisting of operation information indicating a predetermined operation of the motor 220 and status information indicating the status of the hoist 101 in the predetermined operation (information indicating the suspended load mass, tension state, motor temperature, and load swing amount) is collected, and a highly accurate state estimation equation that takes into account individual differences in motor characteristics, etc. is learned. Hereinafter, the data set of the collected operation information and status information is referred to as "learned data." Furthermore, in the operation process, the hoist 101 of this embodiment estimates its state using the operation information and the learned state estimation equation.
[0040] The state information of the learning data is a numerical value or a classification set for learning the state estimation formula, and may be set manually or automatically, for example. In the following, the numerical value and classification set for learning in this way are referred to as "reference values," and the numerical value and classification estimated using the state estimation formula are referred to as "estimated values" to distinguish them.
[0041] The state estimation formula is a formula that takes operation information as input and outputs (calculates) estimated values of the suspended load mass, tension state, motor temperature, and load swing amount. Therefore, four states (estimation items) of the hoisting machine 101 are estimated for one motor 220. In other words, each of the four estimation items shares the operation information that is input.
[0042] The input driving information may be a current sample (driving information), but in this embodiment, it is time-series information of a certain number of samples from the past to the present. If past driving information can be used, the expressive power of the state estimation formula is improved, and for example, a state estimation formula that takes into account the hysteresis of the motor 220 can be obtained. However, when the data capacity is limited due to constraints such as the CPU's computing performance and the storage device's capacity, it is sufficient to acquire at least the driving information after ground-breaking.
[0043] FIG. 6 is a diagram showing an example of a state estimation equation (network 600). FIG. 6 shows a network configuration of multitask estimation in network 600. More specifically, network 600 is a fully connected neural network in which the number of nodes in input layer 610 is N, the number of shared layers 620 is "2", and the number of unique layers 630 branched to each estimation item is "1". Driving information is provided to input layer 610, and the value of each estimation item is calculated. In this case, the state estimation equation between each layer is (Equation 1).
[0044] u = Wx + b (Eq. 1)
[0045] Here, x is an input vector for each layer, and in the input layer 610, it is a vector that concatenates the operation information (hereinafter, referred to as the "operation information vector"). u is a vector that indicates the output of each layer. Note that the u vector is a scalar value in the output layers of the suspended load mass, motor temperature, and load swing amount, and in the output layer of the tension state, it is the probability of each of "tension" and "relaxation". In addition, for example, when estimating the suspended load mass, the estimated value itself may be used, or the ratio of the estimated value of the suspended load mass to the rated load of the product may be used. In the latter case, when u=0.5 for a product with a rated load of 1 t, the estimated value of the suspended load mass is 500 kg. W is a coefficient matrix that indicates the weight for each element of the operation information vector x. b is a vector that indicates the bias.
[0046] Note that the state estimation unit 513 may remove time series noise after state estimation by using a low-pass filter such as a moving average filter as post-processing. The hoist 101 may also perform overload prevention control, life prediction, etc. using estimated values of the suspended load mass and / or motor temperature, detect ground run-out using estimated values of tension state, and perform emergency stop in the event of dangerous swing, load swing reduction control, etc. using estimated values of load swing amount.
[0047] Next, the learning data will be described. The conditions for collecting the learning data (hereinafter referred to as "collection conditions") may include the positional deviation of the suspended load 130 relative to the trolley 113 and the lifting weight pattern during lifting. The lifting weight pattern refers to the hoisting mode such as the high-speed hoisting mode and the low-speed hoisting mode, the hoisting speed, etc. It is desirable that the collected learning data more thoroughly covers the ground lifting patterns expected in the operation of the hoist 101, which enables more accurate state estimation.
[0048] Moreover, it is desirable that the pattern of the suspended load mass when collecting learning data should be a gradual and uniform range from 0 kg to an overload weight (a weight exceeding the rated load of the hoist 101). It is also desirable that the collection conditions include at least one of the maximum load that can be lifted by the hoist 101, the rated load, and no load. This is because the rated load is a representative design point for each product model, the maximum load is an operating point that must be confirmed for product safety, and no load is an operating point that can obtain data that is least affected by noise such as load variation.
[0049] Next, the operation and management of the hoist system 100 will be described.
[0050] Fig. 7 is a diagram showing an example of a flowchart relating to the processing procedure of the hoist 101. The operation of the hoist 101 based on the flowchart of Fig. 7 is as follows. Note that in the pre-shipment process of the hoist 101, processing of steps S701 to S703 is performed, and in the operation process of the hoist 101, processing of step S711 is performed.
[0051] Step S701: A specific computer collects learning data (referred to as "lineup learning data") obtained when a ground cutting test is conducted under conditions (collection conditions) in which the load mass, the positional deviation of the load 130 relative to the trolley 113, and the lifting weight pattern are varied for each model. The lineup learning data is data including operation information (rotation speed, current value, etc.) and ground cutting conditions (load mass, positional deviation, etc.). The specific computer may be a learning device 910, a state estimation device 1310, a cloud server 1610, etc., which will be described later. The ground cutting conditions are conditions input by the operator of the hoisting machine 101, conditions in which some or all of the collection conditions are automatically set, etc., and include state information (reference values) at the time of ground cutting.
[0052] Step S702: A predetermined computer learns a standard state estimation equation for each model using the lineup learning data collected in step S701. For example, the predetermined computer executes optimization learning of the coefficient matrix W and bias b for the network 600 using the backpropagation method.
[0053] Step S703: A specified computer collects learning data (referred to as "individual learning data") obtained when a ground-cutting test is performed during product shipping testing, and re-learns the state estimation formula for each product using the newly collected individual learning data to reflect individual differences in motor characteristics. As a result, the hoist 101 becomes capable of highly accurate state estimation for each product. Note that, if there is little individual difference in motor characteristics for each product, step S703 is not essential.
[0054] The learned state estimation equation for the hoist 101 learned in the pre-shipment process is set so as to be usable by the state estimation unit 513 (for example, stored in the auxiliary storage device 516).
[0055] Step S711: The hoisting machine 101 estimates a state during operation. For example, when an operator operates the hoisting machine 101, the hoisting machine 101 estimates a state from operation information during operation. In the operation of this embodiment, a learned state estimation equation learned in the pre-shipment process is used to estimate the suspended load mass, etc. during operation.
[0056] FIG. 8 is a diagram showing an example of a flowchart relating to the estimation of a state during driving (step S711).
[0057] Step S801: The operator sets the load 130 on the hoist 101.
[0058] Step S802: The operator performs a ground cutting operation via the operation input unit 522, and the motor control unit 511 generates a motor control command in accordance with the operation command input via the operation input unit 522, and outputs the motor control command to the motor 220.
[0059] Step S803: The driving information acquisition unit 512 acquires driving information. More specifically, the driving information acquisition unit 512 acquires driving information from the motor control unit 511 and / or the motor 220, and transmits the acquired driving information to the state estimation unit 513. The driving information includes at least one of information indicating an input value (e.g., an input signal) to the motor 220 and information indicating an output value (e.g., an output signal) from the motor 220.
[0060] Step S804: The state estimation unit 513 performs state estimation (calculates estimated values of the suspended load mass, tension state, motor temperature, and load swing amount) based on the operating information acquired in step S803 and the learned state estimation equation.
[0061] Step S805: The state display unit 521 displays, on the input / output device 527, information indicating the result of the state estimation performed in step S804 (estimated values, a graph indicating the estimated values, values calculated from the estimated values, etc.).
[0062] As described above, the hoisting machine 101 estimates the state based on the state estimation equation in step S804, and notifies the operator of the estimation result via the state display unit 521.
[0063] In this embodiment, in step S703, learning data from a ground test at the time of product shipment test is recorded, and re-learning is performed for each product, but step S703 is not essential. However, by performing step S703, it is possible to estimate the state with high accuracy by reflecting the characteristics of each product.
[0064] Next, the effects of this embodiment will be described.
[0065] Generally, the power output from the motor 220 in the hoist 101 is affected by losses in various components (constituent parts) in the power transmission system, such as the bearings 230, belts, gear chains, brakes, etc. Therefore, in order to estimate the state using a physical model, it is necessary to consider factors such as factors caused by the suspended load mass, factors caused by the components of the power transmission system, and factors caused by individual differences in motor characteristics, separately for the load of each hoist 101.
[0066] On the other hand, to improve the accuracy of state estimation, it is necessary to tune the coefficients formulated for each factor for each individual product, which takes time to set the coefficients.In addition, as the physical model becomes more complex, the calculation load of state estimation also increases, which makes it difficult to achieve, as the cost of the control board required to perform high-precision calculations at high speed tends to increase.
[0067] In contrast, when a state estimation formula based on learning data as described in this embodiment is used, the suspended load mass, tension state, motor temperature, and load swing amount can be estimated with high accuracy by setting the coefficient matrix W and bias b in (Formula 1) to appropriate matrices and vectors. Furthermore, because feature extraction and parameter adjustment are performed automatically through learning, the burden required for investigation can be reduced. In addition, because a complex physical model is not used, the calculation load is reduced, and there are other advantages such as the ability to suppress increases in the cost of the control board.
[0068] The above is an effect that can be obtained even when one state (single state) is estimated, but the effect is further enhanced in the case where four states (multiple states) are estimated from one driving information as shown in this embodiment.
[0069] For example, in the state estimation in this embodiment, a neural network is used, and some of the calculations of the four estimation items are shared in the shared layer 620. This is possible because the four estimation items are items having a parallel connection relationship that can be estimated from the operation information of the same motor 220, or the estimation items are items having a series connection relationship (for example, a dependency relationship).
[0070] To explain more specifically, in the calculation of the tension state, a method of outputting a "tension" judgment when a predetermined suspended load mass is exceeded is possible, so the suspended load mass and the tension state are in a relationship in which they can be estimated from the same operating information. In other words, the suspended load mass and the tension state are in a parallel relationship because the physical background required to estimate each is the same.
[0071] Regarding the motor temperature, the current value changes due to a change in the internal resistance of the motor 220 caused by a change in the temperature of the windings, and this may change the output value of the suspended load mass. Therefore, the suspended load mass and the motor temperature can be considered to be in a dependent relationship (series relationship).
[0072] Furthermore, regarding the amount of load sway, when the suspended load 130 is vibrating, the tension of the rope changes periodically due to the effect of time fluctuations in centrifugal force, etc. In other words, this corresponds to time fluctuations occurring in the estimated value of the suspended load mass (estimated load), so the amplitude of vibration can be estimated from the period of the time fluctuation and the magnitude of the change in the estimated load. Therefore, the amount of load sway and the suspended load mass are in a parallel relationship and can be estimated from the same operational information.
[0073] As described above, by estimating from one state estimation formula the states having a parallel relationship that can be estimated from the same operation information and the states having a serial relationship such as a dependency relationship, it becomes possible to simultaneously estimate many states related to the hoisting machine 101 while keeping the calculation processing and parameter settings to a necessary minimum. Therefore, the more states that are estimated for one motor 220, the more effective the state estimation using learning becomes.
[0074] In this embodiment, it is assumed that a neural network is used for state estimation, but various supervised learning methods can be used. For example, in this embodiment, the suspended load mass, motor temperature, and load swing amount are estimated by regression calculation, and the tension state is estimated by classification calculation, but various learning methods such as simple linear regression, random forest, support vector machine, and logistic regression can be used depending on the format of the desired estimated value.
[0075] Furthermore, in this embodiment, the integrated neural network is treated as one state estimation equation to estimate the suspended load mass, tension state, motor temperature, and load sway amount, but it may also be decomposed, for example, by estimating the suspended load mass and tension state from a first state estimation equation, and estimating the motor temperature and load sway amount from a second state estimation equation having different operating information.
[0076] With the above mechanism, the hoisting machine system 100 can provide a state estimation formula that can estimate with high accuracy such conditions as the suspended load mass, tension state, motor temperature, and load swing amount with simple parameter setting. In particular, the hoisting machine system 100 does not perform re-learning during the operation process, so a learning device is not required for the product after shipment, and the configuration of the operation terminal 120 can be reduced, resulting in a simple system and making it possible to provide an inexpensive hoisting machine 101.
[0077] (II) Second embodiment The second embodiment will be described with reference to Figures 9 to 12. Since the configuration and basic operation of each element in this embodiment are similar to those in the first embodiment, the features of this embodiment will be mainly described.
[0078] 9 is a diagram showing an example of the configuration of a hoist system 900 according to this embodiment. This hoist system 900 includes a hoist 101, an operation terminal 120, and a learning device 910. Information exchange between these devices may be by wired communication or wireless communication.
[0079] The hoisting machine 101 includes a motor 220 connected to a hoist drum 210 that hoists a rope or the like that has a sling attached to a load 130, a motor control unit 511, an operating information acquisition unit 512, and a state estimation unit 513.
[0080] This embodiment differs from the first embodiment in that learning data made up of driving information and state information is collected in an operation process and learning is performed, and therefore there are differences in the driving information acquisition unit 512 and the state estimation unit 513.
[0081] Hereinafter, a flow in which the hoist system 900 estimates a state in the operation process is referred to as a state estimation phase, and a flow in which the hoist system 900 updates a state estimation equation through learning is referred to as a learning phase. The state estimation phase and the learning phase may be clearly separated, or the state estimation phase and the learning phase may be performed in parallel.
[0082] In the state estimation phase, the driving information acquisition unit 512 transmits driving information to the state estimation unit 513, and in the learning phase, transmits driving information to the operation terminal 120 so that learning data can be set.
[0083] In the state estimation phase, the state estimation unit 513 estimates the suspended load mass, tension state, motor temperature, and load swing amount using the state estimation equation and the operation information acquired by the operation information acquisition unit 512. Furthermore, the state estimation unit 513 estimates the deterioration state of the components of the hoist 101, such as the motor 220, bearings 230, gears, belts, chains, brakes, etc., using the latest learning data stored in the learning device 910 or the operation information acquired by the operation information acquisition unit 512, and non-latest learning data (learning data from a past period).
[0084] That is, the state estimation formula in the first embodiment uses operating information at the current time (relevant time) in the operation process, whereas the state estimation formula in the present embodiment uses operation information (learning data) from a past time distant from the relevant time in addition to operation information at the relevant time in the operation process.
[0085] In addition, during the learning phase, the state estimation unit 513 receives the state estimation equation stored in the learning device 910, and the state estimation equation used by the state estimation unit 513 is updated to a state estimation equation suitable for the operating environment of the hoist 101.
[0086] Next, a description will be given of the operation terminal 120. The operation terminal 120 includes a state display unit 521, an operation input unit 522, a learning data condition input unit 921, and a learning data setting unit 922.
[0087] In the state estimation phase, the state display unit 521 displays the estimated value (state related to the hoist 101) calculated by the state estimation unit 513, etc. Furthermore, when the state estimation unit 513 calculates an incorrect estimated value, the state display unit 521 may display a confirmation screen for confirming with the operator that a reference value for a correct state is input and re-learning is performed using learning data including the reference value. Re-learning will be described later. The operation input unit 522 is an interface for the operator to input a value of a motor control command for the hoist 101 in the state estimation phase.
[0088] The learning data condition input unit 921 is an interface for inputting learning data conditions such as the state of the motor 220 or the load 130, the individual identification number of the hoist 101, etc., during the learning phase. Here, the learning data conditions are the operating conditions (e.g., ground-cutting conditions) of the hoist 101 corresponding to the operating information acquired by the operating information acquisition unit 512, and are associated with at least any one of state information (reference values) of the suspended load mass, tension state, motor temperature, and load swing amount. The learning data conditions may also include information such as the model name and individual identification number of the product. The interface of the learning data condition input unit 921 may be used in common with the interface of the operation input unit 522, or may be implemented as a separate interface.
[0089] The learning data setting unit 922 associates the learning data conditions input from the learning data condition input unit 921 with the driving information transmitted from the driving information acquisition unit 512 to set them as one set of learning data, and transmits them to the learning device 910.
[0090] In this embodiment, the status display unit 521, the operation input unit 522, the learning data condition input unit 921, and the learning data setting unit 922 are implemented in the operation terminal 120, but may also be implemented in the hoist 101.
[0091] Next, the learning device 910 will be described. The learning device 910 includes a learning data storage unit 911, a learning unit 912, and a learning result storage unit 913. The learning device 910 also includes hardware such as a calculation unit 914 such as a CPU, a main storage unit 915 such as a semiconductor memory, an auxiliary storage unit 916 such as a hard disk, and a communication unit 917. Well-known operations of each piece of hardware will be omitted as appropriate in the following description.
[0092] The functions of the learning device 910 (learning data storage unit 911, learning unit 912, learning result storage unit 913, etc.) may be realized, for example, by the arithmetic device 914 reading a program stored in the auxiliary storage device 916 into the main storage device 915 and executing it (software), or may be realized by hardware such as a dedicated circuit, or may be realized by a combination of software and hardware. Note that one function of the learning device 910 may be divided into multiple functions, or multiple functions may be combined into one function. Also, some of the functions of the learning device 910 may be provided as a separate function, or may be included in another function. Also, some of the functions of the learning device 910 may be realized by another computer that can communicate with the learning device 910.
[0093] The learning data storage unit 911 stores the learning data set by the learning data setting unit 922 of the operation terminal 120 in the auxiliary storage device 916. The stored learning data is classified into three types.
[0094] The first is lineup training data, which is training data collected under different collection conditions for each model, and is used to train a standard state estimation formula for each model in the first pre-shipment process.
[0095] The second type is individual learning data, which is learning data for each product collected during quality assurance testing (e.g., ground cutting tests) at the time of product shipment testing, and is used for re-learning to reflect individual differences in motor characteristics during the pre-shipment process.
[0096] The third type is learning data under the operating environment, which is learning data in which the operator corrects the estimated values calculated by the state estimation unit 513 after the product is shipped, and is used for re-learning to reflect the characteristics of the operating environment in the operating process.
[0097] The learning unit 912 reads the learning data from the learning data accumulation unit 911 and learns the state estimation equation. The lineup learning data is used in learning the standard state estimation equation for each model in the first step of the shipping process. Moreover, in re-learning that reflects individual differences in motor characteristics in the pre-shipment process, learning data in which individual learning data is added to the lineup learning data is used. Moreover, in re-learning in the operation process, learning data in which learning data under the operation environment is added to the lineup learning data or the individual learning data is used.
[0098] The learning result accumulation unit 913 accumulates the state estimation equation learned by the learning unit 912 in the auxiliary storage device 916, and transmits the state estimation equation as appropriate to the state estimation unit 513. Note that the learning result accumulation unit 913 may accumulate a standard state estimation equation for each model, and further, state estimation equations for each individual device and each operating environment.
[0099] These state estimation equations may be able to be referenced as appropriate by specifying the product model name, individual identification number, operating environment (for example, fastener, load mass, volume of the load 130, material of the load 130, quantity of the load 130, shape of the load 130, specificity of the load 130 (flexibility, liquidity, temperature, harmfulness, fragility, etc.), method of tying the rope, number of ropes to be tyed, hoisting angle), etc. In this way, by providing a state estimation equation for each operating environment, the operator of the hoist 101 can easily use different state estimation equations according to the operating environment.
[0100] Next, the learning (for example, learning calculation) performed by the learning unit 912 will be described in detail. The learning calculation refers to a process of adjusting the coefficient matrix W and the bias b so that the estimated value u output when the driving information vector x is input in the state estimation formula given by (Formula 1) in the first embodiment is an estimated value with less error.
[0101] The flow of the learning calculation will be described. First, the learning unit 912 randomly extracts some learning data from a part or all of the learning data (referred to as "learning data to be learned") stored in the learning data storage unit 911. Then, the learning unit 912 inputs the extracted driving information vector x into (Equation 1) into which, for example, the initial value of the randomly determined coefficient matrix W and the initial value of the bias b are substituted, and calculates an estimated value u according to (Equation 1).
[0102] The learning unit 912 uses the calculated estimated value u to calculate an error e between the reference value u' stored in the learning data, and calculates a weighted error es by weighting and summing up the errors e of each estimation item according to the priority. Since the calculated weighted errors es are the same as the number of extracted learning data, the learning unit 912 sums them up to calculate a weighted total error E.
[0103] Then, the learning unit 912 slightly changes the coefficient matrix W and the bias b so as to reduce the weighted sum error E. As a method for slightly changing the coefficient matrix W and the bias b here, for example, a gradient descent method or the like can be used.
[0104] The learning unit 912 repeats the same process for the remaining learning data from the learning data to be learned, sequentially and randomly extracting a similar number of learning data, and when the process is completed for all of the learning data to be learned, one cycle of the learning process is completed.
[0105] By repeating the above procedure, the weighted total error E gradually decreases, and the learning unit 912 continues to update the state estimation formula until a learning end condition is met. The learning end condition is a condition for completing learning of the state estimation formula. For example, the learning end condition is when the number of times the learning cycle is repeated exceeds a predetermined number, when the weighted total error E reaches a predetermined value or less, etc. The estimated value u output by the state estimation formula obtained in this way becomes an optimal estimated value for the driving information vector x acquired at that time point.
[0106] Moreover, the act of the learning unit 912 performing learning calculations by providing the coefficient matrix Wa and bias ba of a learned state estimation equation as the initial values of the coefficient matrix W and bias b is referred to as "relearning." The learning unit 912 can update the state estimation equation to one that fits the additional learning data by adding new learning data to the learning data and performing learning calculations.
[0107] In this embodiment, the learning unit 912 uses the state estimation equation created using the lineup learning data as an initial value and performs re-learning using the individual learning data, thereby updating the state estimation equation to one that reflects individual differences in the motor characteristics of the product. In addition, the learning unit 912 performs re-learning using the learning data in the operating environment, thereby updating the state estimation equation to one that reflects the characteristics of the operating environment. This enables more accurate state estimation that reflects the characteristics of each product and the characteristics of the operating environment in which the product is used.
[0108] The state estimation equation, which includes the learned coefficient matrix W and bias b calculated by the above procedure, is stored in the learning result storage unit 913 and transmitted when the state estimation unit 513 is updated, enabling the hoisting machine 101 to perform highly accurate state estimation.
[0109] In addition, this embodiment differs from the first embodiment in that the state estimation unit 513 estimates the deterioration state of the components of the power transmission system, such as the motor 220 and the bearings 230, and therefore the calculation processing related to the estimation of the deterioration state will be described.
[0110] In the state estimation unit 513 of this embodiment, in addition to the time-series operation information (time-series information) at the time of operation, learning data of an arbitrary time is read out from the learning data accumulation unit 911. Here, the arbitrary time can be set by the operator through the operation input unit 522, and for example, the initial time at the start of operation (at the time of shipment), immediately after regular maintenance (at the time of inspection), etc. can be selected.
[0111] In the state estimation unit 513, the suspended load mass, tension state, motor temperature, and load swing amount are calculated from the current operation information (time series information), the suspended load mass is estimated when past operation information acquired from learning data of past periods is used, and the difference from the suspended load mass estimated from the current operation information is calculated. The relationship between the difference between the current and past suspended load masses and deterioration over time is described in advance as an equation inside the state estimation formula, the deterioration degree is calculated from the difference in the suspended load mass, and the deterioration degree is transmitted from the state estimation unit 513 to the state display unit 521. The deterioration degree here may be a display value for determining whether maintenance is required or whether parts need to be replaced, or may be a numerical value for the lifespan of the components of the hoist 101. When calculating the deterioration degree, the state estimation unit 513 may refer to a database stored in the learning result storage unit 913 or the learning data storage unit 911.
[0112] With the above mechanism, the operator can use past operating information to know the timing of maintenance of the hoisting machine 101 and the lifespan of the components of the hoisting machine 101, etc., and can therefore prevent the production line from being stopped due to a sudden breakdown of the hoisting machine 101, and prevent the need for maintenance during busy periods, etc.
[0113] In this embodiment, the deviation between the past and present estimated values of the suspended load mass is evaluated, but it is also possible to simultaneously use estimated values of other estimation items such as tension state, motor temperature, load swing amount, etc. Also, instead of using a specific output value, a method using internal calculations such as inputting both current and past operating information (learning data) to the input layer 610 of the neural network, extracting respective feature amounts during the calculation process, and outputting the deterioration degree in the output layer may be used.
[0114] Also, there is a method of estimating the deterioration level using a plurality of the state estimation formulas stored by the learning result storage unit 913 and the learning data stored by the learning data storage unit 911, and selecting the one with higher safety from the two deterioration levels. Therefore, similarly to the first embodiment, it is possible to use multiple supervised learning methods, not limited to neural networks, and further to use ensemble learning combining multiple learning methods.
[0115] 10 is a diagram showing an example of a flowchart relating to the processing procedure of the hoist system 900. The operation of the hoist system 900 based on the flowchart is as follows. Note that in the pre-shipment process of the hoist 101, processing from step S701 to step S703 is performed, and in the operation process of the hoist 101, processing from step S1001 to step S1005 is performed. Since steps S701 to step S703 are the same as in the first embodiment, steps S1001 to step S1005 will be described.
[0116] Step S1001: When an operator operates the hoisting machine 101, the hoisting machine 101 estimates a state during operation. More specifically, the hoisting machine 101 estimates (calculates estimated values) a suspended load mass, a tension state, a motor temperature, and a load swing amount by using operation information during operation and a state estimation equation.
[0117] Step S1002: If the operator determines that the estimated value calculated in step S1001 is incorrect, the operator corrects the estimated value. In step S1002, the operation terminal 120 asks the operator via the status display unit 521 whether correction is necessary. If the operator determines that correction is necessary, the operation terminal 120 moves the process to step S1003. If the operator determines that correction is not necessary, the operation terminal 120 moves the process to step S1004.
[0118] Step S1003: The learning data setting unit 922 associates the information corrected by the operator (reference value) with the driving information used for state estimation, and sets it as learning data (referred to as "learning data for re-learning") in the operating environment. Then, the learning data setting unit 922 transmits the learning data for re-learning to the learning data storage unit 911.
[0119] Step S1004: The learning device 910 determines whether or not to perform re-learning. The learning device 910 determines to perform re-learning when it receives an instruction from an operator, when a certain amount of learning data for re-learning has been accumulated, when a certain degree of degradation has been exceeded, etc. If the learning device 910 determines to perform re-learning (in the case of "YES"), it proceeds to step S1005, and if it determines not to perform re-learning (in the case of "NO"), it ends the process.
[0120] Step S1005: The learning unit 912 performs re-learning using the learning data for re-learning collected in step S1003, and transmits a state estimation equation reflecting the characteristics of the operating environment to the learning result accumulation unit 913. The state estimation equation is transmitted to the state estimation unit 513, and state estimation is performed using the new state estimation equation from the next operation.
[0121] This allows the state to be estimated with high accuracy, reflecting the characteristics of the operating environment. Note that, if the motor characteristics do not change much due to the characteristics of the operating environment, step S1005 is not essential.
[0122] In this embodiment, the learning unit 912 uses the learning data to learn and calculate the state estimation equation in steps S702, S703, and S1005, so that the state estimation equation can be updated according to the tendency of the learning data. Therefore, individual differences in the motor characteristics of the product, characteristics of the operating environment, and even aging deterioration can be reflected in the state estimation equation, and as a result, a more accurate state can be estimated. Furthermore, even without building physical models corresponding to different models of hoisting machines 101, the hoisting machine system 900 can estimate the state by collecting learning data using the same processing.
[0123] FIG. 11 is a diagram showing an example of a flowchart relating to detailed procedures of steps S701 and S703 for collecting learning data by a ground-cutting test in the pre-shipment process.
[0124] Step S1101: The manufacturer determines the conditions for collecting learning data in advance. The conditions for collecting learning data are the conditions for the load mass in the ground-lifting test, the positional deviation of the load 130 relative to the trolley 113, and the combination of lifting weight patterns, and it is desirable to set them thoroughly so as to cover all the conditions expected when the hoist 101 is operated. Then, the manufacturer sets the load 130 on the hoist 101 according to the conditions for collecting learning data.
[0125] Step S1102: The manufacturer inputs the ground lifting operation from the operation input unit 522, and the hoisting machine 101 records the operation information (time-series information) during the ground lifting operation. The time-series information is preferably information from the start of ground lifting to when the suspended load 130 becomes stable.
[0126] Step S1103: The manufacturer inputs the learning data conditions (ground cutting conditions) from the learning data condition input unit 921. In step S1103, some or all of the predetermined collection conditions of the learning data are automatically set as the learning data conditions, and if there is a missing condition in the set learning data conditions, the manufacturer may add that condition, or if there is an unnecessary condition in the set learning data conditions, delete that condition.
[0127] Step S1104: The learning data setting unit 922 associates the driving information (time-series information) recorded in step S1102 with the learning data conditions input in step S1103, and sets them as learning data.
[0128] Step S1105: The learning data setting unit 922 transmits the learning data set in step S1104 to the learning data storage unit 911. The learning data storage unit 911 stores the received learning data.
[0129] As described above, in this embodiment, the learning data is collected by performing a ground-cutting test, and the state estimation formula is learned and calculated using the collected learning data, thereby enabling accurate state estimation. In addition, the learning data collection work is not subject to personal judgment, and has the advantage that it can be performed without advanced knowledge of products such as the motor 220 of the hoist 101.
[0130] FIG. 12 is a diagram showing an example of a flowchart relating to a detailed procedure of the processes (steps S702, S703, and S1005) relating to the learning calculation of the state estimation formula.
[0131] Step S 1201 : The learning unit 912 reads the learning data from the learning data accumulation unit 911 .
[0132] Step S1202: The learning unit 912 performs learning calculations on the state estimation equation using the learning data read in step S1201.
[0133] Step S1203: The learning result storage unit 913 stores the state estimation equation (learning result) learned and calculated in step S1202.
[0134] Step S1204: The learning result accumulation unit 913 transmits the state estimation equation learned and calculated in step S1202 to the state estimation unit 513.
[0135] Through the above procedure, the state estimation equation is learned and calculated so as to reflect the learning data collected in step S1003 and the like.
[0136] As described above, this embodiment does not require any additional sensors, and by learning using operating information and status information (reference values) at the time of ground cutting for each model, product, and operating environment, it is possible to perform highly accurate status estimation that takes into account individual differences in motor characteristics, characteristics of the operating environment, and even the deterioration state of the components of the hoisting machine 101.
[0137] (III) Third embodiment The third embodiment will be described with reference to Fig. 13 and Fig. 14. In this embodiment, a case will be described in which a state estimating device 1310 is attached to a hoisting machine 101 that has already been installed or a hoisting machine 101 that has already been purchased.
[0138] 13 is a diagram showing an example of the configuration of a hoist system 1300 in this embodiment. The hoist 101 is equipped with a motor 220 that hoists a rope or the like slung around a suspended load 130. Note that a description of the same configuration as in the first embodiment will be omitted.
[0139] The state estimation device 1310 includes a driving information acquisition unit 512, a state estimation unit 513, a state display unit 521, a learning data accumulation unit 911, a learning unit 912, a learning result accumulation unit 913, a learning data condition input unit 921, and a learning data setting unit 922.
[0140] The state estimation device 1310 includes hardware such as a calculation device 1311 such as a CPU, a main storage device 1312 such as a semiconductor memory, an auxiliary storage device 1313 such as a hard disk, a communication device 1314, and an input / output device 1315. Well-known operations of each piece of hardware will be omitted as appropriate in the following description.
[0141] The functions of the state estimation device 1310 (the driving information acquisition unit 512, the state estimation unit 513, the state display unit 521, the learning data accumulation unit 911, the learning unit 912, the learning result accumulation unit 913, the learning data condition input unit 921, the learning data setting unit 922, etc.) may be realized, for example, by the arithmetic device 1311 reading a program stored in the auxiliary storage device 1313 into the main storage device 1312 and executing it (software), or may be realized by hardware such as a dedicated circuit, or may be realized by a combination of software and hardware. Note that one function of the state estimation device 1310 may be divided into multiple functions, or multiple functions may be integrated into one function. Also, some of the functions of the state estimation device 1310 may be provided as a separate function, or may be included in another function. Also, some of the functions of the state estimation device 1310 may be realized by another computer that can communicate with the state estimation device 1310.
[0142] The operation information acquisition unit 512 acquires operation information of the motor 220 of the hoist 101. The acquired operation information is information indicating an input value to the motor 220 and information indicating an output value from the motor 220. The state estimation unit 513 calculates estimated values of states such as the suspended load mass, tension state, motor temperature, and load swing amount from the operation information acquired by the operation information acquisition unit 512 using a state estimation formula. The state display unit 521 displays the estimated values calculated by the state estimation unit 513. The learning data condition input unit 921 is an interface for inputting learning data conditions, and performs processing related to the input / output device 1315 such as a touch panel, a keyboard, and a mouse. The learning data conditions input by the learning data condition input unit 921 are transmitted to the learning data setting unit 922.
[0143] The learning data setting unit 922 associates the learning data conditions input by the learning data condition input unit 921 with the driving information acquired by the driving information acquisition unit 512 to set one set of learning data, and transmits the set learning data to the learning data accumulation unit 911. The learning data accumulation unit 911 accumulates the learning data set by the learning data setting unit 922. The learning unit 912 reads the learning data from the learning data accumulation unit 911, and learns and calculates a state estimation equation.
[0144] The learning result accumulation unit 913 accumulates the state estimation equation learned and calculated by the learning unit 912. Then, the learning result accumulation unit 913 appropriately transmits the state estimation equation to the state estimation unit 513 at the timing when the state estimation equation is updated, etc.
[0145] Next, the processing procedure performed by the state estimation device 1310 will be described.
[0146] FIG. 14 is a diagram showing an example of a flowchart relating to a processing procedure of the hoist system 1300.
[0147] Step S1401: The installer attaches the state estimation device 1310 to the hoisting machine 101.
[0148] Step S1402: The installer conducts a ground-cutting test, and the learning data setting unit 922 associates the operating information before and after ground-cutting with the ground-cutting conditions (suspended load mass, tension state, motor temperature, and status information on load swing amount) and sets them as learning data, and the learning data storage unit 911 stores the learning data.
[0149] Step S1403: The learning unit 912 learns and calculates the state estimation equation, and the learning result storage unit 913 stores the state estimation equation learned and calculated by the learning unit 912.
[0150] Step S1404: The state estimation unit 513 updates the state estimation equation and performs state estimation during driving.
[0151] Note that steps S1402, S1403, and S1404 are similar to steps S701, S702, and S711, respectively. Furthermore, following step S1404, the processes of steps S1002 to S1005 may be performed.
[0152] In this embodiment, by attaching a state estimation device 1310 to an existing hoisting machine 101, even if the hoisting machine 101 does not have a state estimation function at the time of product shipment, state estimation can be performed later.
[0153] (IV) Fourth embodiment The fourth embodiment will be described with reference to FIG.
[0154] 15 is a diagram showing an example of the configuration of a hoist system 1500 of this embodiment. In this embodiment, a plurality of hoists 101 (three in this example) exist for one operation terminal 120.
[0155] The elements of the input / output system and the estimation items built into each of the hoisting machines 101-1 (hoisting machine A), 101-2 (hoisting machine B), 101-3 (hoisting machine C), the operation terminal 120, and the learning device 910 are the same as those in the second embodiment. However, a difference is that the path for transmitting the state estimation equation estimated by the learning device 910 to the hoisting machine 101 passes through the operation terminal 120. In other words, in the present embodiment, in each of the hoisting machines 101, the suspended load mass, motor temperature, tension state, load swing amount, and deterioration state of the motor 220 and the power transmission system are estimated from the operation information of one motor 220, so that the relationship between the number of motors 220 and the number of estimation items is the same as that in the second embodiment.
[0156] The learning data collected by each hoist 101 is collected and stored in a single learning device 910 via a single operation terminal 120. Then, a state estimation equation learned and calculated in the learning device 910 based on the learning data is stored in a learning result storage unit 913. The state estimation equation stored in the learning result storage unit 913 is appropriately transmitted to a state estimation unit 513 built into the hoist 101 via the operation terminal 120 at the timing of maintenance or the like.
[0157] According to the hoist system 1500, the burden on the operator can be reduced because multiple hoists 101 can be operated with the same operation terminal 120 and learning device 910. In addition, operation can be achieved by preparing one learning device 910, which is generally expensive.
[0158] Furthermore, by storing operation information (learning data) of a plurality of hoists 101 in a single learning device 910, it is possible to utilize the operation information (learning data) of the hoists 101. For example, when hoist A is used frequently and hoists B and C are used infrequently, a lot of learning data of hoist A is accumulated, and therefore changes in the state estimation formula and learning data due to aging degradation are quickly accumulated.
[0159] As a result, when learning data for hoist B and hoist C has been accumulated, it is possible to estimate the degree of deterioration by comparing it with the progress of hoist A, which is used in the same operating environment, thereby improving the accuracy of deterioration estimation.
[0160] (V) Fifth embodiment The fifth embodiment will be described with reference to FIG.
[0161] 16 is a diagram showing an example of the configuration of a hoisting machine system 1600 of this embodiment. In the hoisting machine system 1600, the learning devices in the second to fourth embodiments are provided on a cloud server 1610. In this embodiment, it is assumed that different operation terminals 120 are connected to one cloud server 1610, and hoists 101 corresponding to each operation terminal 120 are in operation.
[0162] Cloud server 1610 may be a server device owned by the manufacturer of the product, a server device owned by the purchaser of the product, or a server device owned by a third party that provides cloud services.
[0163] The learning data acquired by each hoist 101 is labeled with the individual identification number, model, etc. of the hoist 101 from which the learning data was acquired, and is stored on the cloud server 1610 via the operation terminal 120.
[0164] Generally, computers for learning are required to have high performance, so they tend to be expensive. In addition, they require preparation of the installation environment and regular maintenance, which incurs management costs.
[0165] In contrast, by using a cloud server 1610 managed by a specialized department or company as in this embodiment, operators no longer need to install, operate, and maintain each learning device 910, thereby reducing the operational burden.
[0166] Furthermore, by storing various learning data for different models and operating environments in the same server device, the vast amount of learning data stored in the same server device can be used when estimating aging deterioration and adapting to changes in the operating environment in the second and fourth embodiments, etc. This makes it possible to further improve the accuracy of diagnosing aging deterioration and adapting to the operating environment.
[0167] As described above, the first to fifth embodiments can provide a highly accurate state estimation formula that takes into account the influence of multiple factors for state estimation, including the determination of the tension state, by a simple parameter setting operation using learning. The introduction effect is particularly high when estimating the state of estimation items in which the operating information used for state estimation has a parallel relationship and / or an estimation item in which the operating information has a serial relationship. In addition, by introducing a mechanism for accumulating learning data and updating the state estimation formula, a state estimation formula that takes into account aging deterioration can be provided, and further, by sharing learning data of multiple hoists, the estimation accuracy of aging deterioration and the accuracy of adaptation to various operating environments can be further improved.
[0168] The above-mentioned first to fifth embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those having all the configurations described. In addition, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is also possible to add, delete, or replace a part of the configuration of each embodiment with another configuration.
[0169] (VI) Notes The above-described embodiment includes, for example, the following contents.
[0170] In the above-described embodiment, the present invention is described as being applied to a hoist system, but the present invention is not limited to this and can be widely applied to various other systems, devices, methods, and programs.
[0171] In the above-described embodiment, a part or all of the program may be installed from a program source into a device such as a computer that realizes the state estimation function. The program source may be, for example, a program distribution server device connected via a network or a computer-readable recording medium (e.g., a non-transitory recording medium). In the above description, two or more programs may be realized as one program, or one program may be realized as two or more programs.
[0172] In the above-described embodiment, the output of information is not limited to display on a display. The output of information may be audio output from a speaker, output to a file, printing on a paper medium or the like by a printing device, projection on a screen or the like by a projector, or other modes.
[0173] In addition, in the above description, information such as programs, tables, files, etc. that realize each function can be placed in a storage device such as a memory, a hard disk, or an SSD (Solid State Drive), or in a recording medium such as an IC card, an SD card, or a DVD.
[0174] The above-described embodiment has the following characteristic configurations, for example.
[0175] (1) A hoist (e.g., hoist 101) that has a motor (e.g., motor 220) and controls the motor to hoist and lower a load is provided with an operation information acquisition unit (e.g., operation information acquisition unit 512) that acquires operation information of the motor, and a state estimation unit (e.g., state estimation unit 513) that estimates a state related to the hoist based on the operation information acquired by the operation information acquisition unit and a state estimation equation (e.g., network 600) that has been learned to be able to estimate a state related to the hoist.
[0176] In the above configuration, the state of the hoist can be estimated by using the motor operation information and the state estimation formula without constructing a physical model. In addition, since a complex physical model is not used in estimating the state of the hoist, for example, the calculation load is reduced and an increase in the cost of the control board can be suppressed.
[0177] The state estimation formula may be learned using lineup learning data, or may be learned using lineup learning data and individual learning data.
[0178] (2) The hoisting machine is equipped with a rope to which a load can be attached, and the state estimation unit estimates at least one of the following states related to the hoisting machine: the mass of the load attached to the rope, the acceleration of the load, the tension of the rope, the rotational load of the motor, the tension state of the rope, the temperature of the motor, and the amount of load sway of the load.
[0179] According to the above configuration, at least one of the mass of the suspended load, the acceleration of the suspended load, the tension of the rope, the rotational load of the motor, the tension state of the rope, the temperature of the motor, and the amount of load sway of the suspended load can be estimated.
[0180] (3) The state estimation unit estimates a predetermined state (e.g., suspended load mass), a state that is in a parallel relationship with the predetermined state (e.g., tension state, load swing amount), and / or a state that is in a series relationship with the predetermined state (e.g., motor temperature) as a state related to the hoisting machine.
[0181] In the above configuration, a predetermined state and a state that has a parallel relationship with the predetermined state, such as a state that shares a common physical background with the predetermined state, and / or a state that has a serial relationship with the predetermined state, such as a state that has a dependency relationship with the predetermined state, are estimated from one piece of driving information. This makes it possible to efficiently estimate a state, for example, without the need to set up a state estimation formula for each state, prepare driving information for each state, or calculate an estimated value for each state.
[0182] (4) The hoist is equipped with an output unit (e.g., a state estimation unit 513) that outputs (e.g., transmits to the operation terminal 120) information indicating the results estimated by the state estimation unit (estimated values, graphs showing the estimated values, values calculated from the estimated values, etc.).
[0183] According to the above configuration, information indicating the state of the hoist is output. For example, the hoist can display the information so that an operator can check for safety, can use the information for other estimations (estimation of deterioration level, life expectancy prediction, etc.), and can use the information for various controls (overload prevention control, emergency stop, etc.).
[0184] (5) The operation information includes both information indicating an input value to the motor and information indicating an output value from the motor.
[0185] In the above configuration, the state of the hoist is estimated using both information indicating the input value to the motor and information indicating the output value from the motor, so that the accuracy of the estimation can be improved compared to, for example, a case in which only information indicating the input value or information indicating the output value is used.
[0186] (6) The operating information includes at least one of the following information: a voltage value applied to the motor, a command frequency to a control unit (inverter) of the motor, the rotation speed of the motor calculated from the sensor value of an encoder provided for the motor, a drive current value input to the motor, an excitation current value in vector control of the motor, a torque current value in vector control of the motor, an input torque value to the motor, a command speed input to the motor, and a slip value of the rotation speed of the motor relative to the command speed input to the motor.
[0187] (7) The operational information used by the state estimation unit to estimate the state of the hoist is time-series information for a predetermined period of time.
[0188] According to the above configuration, driving information for a specified period (e.g., several seconds) from a certain point in time to another certain point in time can be used, so that the accuracy of estimation can be improved compared to, for example, a case where driving information at a certain point in time is used.
[0189] (8) The hoist system (hoist system 900) includes a hoist (e.g., hoist 101) that includes a motor (e.g., motor 220) and controls the motor to hoist and lower a load, an operation terminal (e.g., operation terminal 120) for operating the hoist, and a learning device (e.g., learning device 910) that learns a state estimation equation used to estimate a state related to the hoist. The learning device includes a learning data storage unit (e.g., learning data storage unit 911) that stores learning data in which operation information of the motor and information indicating a state related to the hoist (learning data condition, reference value, state information, etc.) are associated with each other, a learning unit (e.g., learning unit 912) that learns a state estimation equation using the learning data stored by the learning data storage unit, and a learning result storage unit (e.g., learning result storage unit 913) that stores the state estimation equation learned by the learning unit. The hoist includes an operation information acquisition unit (e.g., operation information acquisition unit 512) that acquires operation information of the motor, and a state estimation unit (e.g., state estimation unit 513) that estimates a state related to the hoist using the operation information acquired by the operation information acquisition unit and a state estimation formula stored by the learning result storage unit. The operation terminal includes a state display unit (e.g., state display unit 521) that displays information indicating the result estimated by the state estimation unit (estimated values, graphs indicating estimated values, values calculated from estimated values, etc.).
[0190] In the above configuration, for example, by learning the state estimation equation using operation information of the motor of the hoist, it is possible to obtain a state estimation equation that reflects the characteristics of the hoist.
[0191] (9) The operation terminal includes a learning data setting unit (e.g., learning data setting unit 922) that, when information indicating the state of the hoist estimated by the state estimation unit based on specified operating information is corrected, associates the corrected information (e.g., a reference value) with the specified operating information and transmits it to the learning device as learning data for re-learning, and the learning data storage unit stores the learning data for re-learning transmitted from the learning data setting unit.
[0192] In the above configuration, for example, by re-learning the state estimation equation using the learning data for re-learning, it is possible to obtain a state estimation equation that reflects the characteristics of the operating environment.
[0193] (10) The state estimation unit uses the operating information acquired by the operating information acquisition unit, the learning data stored by the learning data storage unit, and the state estimation formula stored by the learning result storage unit to estimate at least one of the deterioration degree of the motor and the deterioration degree of components connected to the power transmission system of the motor as a state related to the hoisting machine.
[0194] In the above configuration, the deterioration degree of the motor and / or the deterioration degree of the components connected to the motor's power transmission system are estimated, so that, for example, an operator can grasp the current and future state of the hoist. In this case, the operator can know the timing of maintenance of the hoist, the life of the hoist, etc., and therefore can prevent the stop of the production line due to a sudden breakdown of the hoist, the implementation of maintenance during busy periods, etc.
[0195] (11) A plurality of the hoists are provided, and the learning data storage unit stores learning data for each of the plurality of hoists (see, for example, FIG. 15).
[0196] In the above configuration, since the learning data including the operation information of the multiple hoists is accumulated by the learning data accumulation unit, for example, the operation information of the mutual hoists can be utilized to estimate the deterioration degree of each hoist. According to the above configuration, for example, it is possible to calculate the deterioration degree in comparison with the motor and the components connected to the motor's power transmission system used in the same operating environment. This makes it possible to more accurately estimate the deterioration degree of the hoist motor and the components connected to the motor's power transmission system.
[0197] (12) A state estimation device (e.g., state estimation device 1310) that includes a motor (e.g., motor 220) and estimates a state related to a hoisting machine that controls the motor to hoist and lower a load includes an operation information acquisition unit (e.g., operation information acquisition unit 512) that acquires operation information of the motor, a learning data storage unit (e.g., learning data storage unit 911) that accumulates learning data in which the operation information acquired by the operation information acquisition unit corresponds to information indicating a state related to the hoisting machine (learning data conditions, reference values, state information, etc.), a learning unit (e.g., learning unit 912) that learns a state estimation formula using the learning data accumulated by the learning data storage unit, a learning result accumulation unit (e.g., learning result accumulation unit 913) that accumulates the state estimation formula learned by the learning unit, and a state estimation unit (e.g., state estimation unit 513) that estimates a state related to the hoisting machine using the operation information acquired by the operation information acquisition unit and the state estimation formula accumulated in the learning result accumulation unit.
[0198] In the above configuration, for example, by attaching a state estimation device to an existing hoist, state estimation is possible even for a hoist that is not equipped with a state estimation function at the time of shipment of the hoist.
[0199] (13) The state estimation device includes a learning data setting unit (e.g., learning data setting unit 922) that, when the motor is controlled according to specified operating conditions of the hoisting machine (e.g., ground cutting conditions) during a test performed when the hoisting machine is shipped or inspected, associates the operating information of the motor acquired by the operating information acquisition unit with the specified operating conditions and sets them as learning data, and the learning data storage unit stores the learning data set by the learning data setting unit.
[0200] In the above configuration, the state estimation equation is learned using operating information of the hoist motor obtained, for example, at the time of shipment of the hoist or during inspection of the hoist, so that a state estimation equation that reflects the characteristics of the hoist or a state estimation equation that reflects the characteristics of the operating environment can be obtained.
[0201] (14) The specified operating conditions include at least one of a condition under which the hoist is at maximum load, a condition under which the hoist is at rated load, and a condition under which the hoist is at no load with no load attached to the hoist.
[0202] The maximum load is an operating point that must be checked for the safety of the hoist, the rated load is a representative design point for each model of hoist, and no load is an operating point that is least affected by noise such as load variation, so that the above configuration makes it possible to appropriately collect learning data.
[0203] Furthermore, the above-described configurations may be modified, rearranged, combined, or omitted as appropriate without departing from the spirit and scope of the present invention.
[0204] It should be understood that items listed in the format "at least one of A, B, and C" can mean (A), (B), (C), (A and B), (A and C), (B and C), or (A, B, and C). Similarly, items listed in the format "at least one of A, B, or C" can mean (A), (B), (C), (A and B), (A and C), (B and C), or (A, B, and C). [Explanation of symbols]
[0205] 100...hoisting machine system, 101...hoisting machine, 120...terminal device.
Claims
1. A hoisting machine including a motor and controlling the motor to hoist and lower a load, An operation information acquisition unit that acquires operation information of the motor; A state estimation unit that estimates a state related to the hoisting machine based on the operation information acquired by the operation information acquisition unit and a state estimation equation that has been learned to be able to estimate a state related to the hoisting machine; Equipped with The operation information includes both information indicating an input value to the motor and information indicating an output value from the motor. Hoisting machine.
2. The hoisting machine includes a rope to which a load can be attached, The state estimation unit estimates at least one of a mass of a load attached to the rope, an acceleration of the load, a tension of the rope, a rotation load of the motor, a tension state of the rope, a temperature of the motor, and a load swing amount of the load as a state related to the hoisting machine. A hoist as claimed in claim 1.
3. The state estimation unit estimates, as a state related to the hoist, a predetermined state, a state in a parallel relationship with the predetermined state, and / or a state in a series relationship with the predetermined state. A hoist as claimed in claim 1.
4. an output unit that outputs information indicating a result of the estimation by the state estimation unit; A hoist as claimed in claim 1.
5. The operation information includes at least one of the following information: a voltage value applied to the motor, a command frequency to a control unit (inverter) of the motor, the rotation speed of the motor calculated from a sensor value of an encoder provided for the motor, a drive current value input to the motor, an excitation current value in vector control of the motor, a torque current value in vector control of the motor, an input torque value to the motor, a command speed to be input to the motor, and a slip value of the rotation speed of the motor relative to the command speed to be input to the motor. A hoist as claimed in claim 1.
6. The operating information used by the state estimation unit to estimate the state of the hoist is time-series information for a predetermined period of time. A hoist as claimed in claim 1.
7. A hoisting machine system including a motor, the hoisting machine controlling the motor to hoist and lower a load, an operation terminal for operating the hoisting machine, and a learning device for learning a state estimation equation used to estimate a state related to the hoisting machine, The learning device includes: A learning data storage unit that stores learning data in which operation information of the motor and information indicating a state related to the hoist are associated with each other; a learning unit that learns a state estimation equation by using learning data stored in the learning data storage unit; a learning result storage unit that stores the state estimation equation learned by the learning unit, The hoisting machine includes: An operation information acquisition unit that acquires operation information of the motor; A state estimation unit that estimates a state related to the hoisting machine using the operating information acquired by the operating information acquisition unit and a state estimation formula stored by the learning result storage unit, the operation terminal includes a state display unit that displays information indicating a result of the estimation by the state estimation unit, The operation information includes both information indicating an input value to the motor and information indicating an output value from the motor. Hoisting machine system.
8. the operation terminal includes a learning data setting unit that, when information indicating a state related to the hoist estimated by the state estimation unit based on predetermined operating information is corrected, associates the corrected information with the predetermined operating information and transmits the corrected information to the learning device as learning data for re-learning, The learning data storage unit stores the learning data for re-learning transmitted from the learning data setting unit. The hoist system of claim 7.
9. The state estimation unit estimates, as a state related to the hoisting machine, at least one of a deterioration degree of the motor and a deterioration degree of a component connected to a power transmission system of the motor, using the driving information acquired by the driving information acquisition unit, the learning data accumulated by the learning data accumulation unit, and the state estimation equation accumulated by the learning result accumulation unit. The hoist system of claim 8.
10. The hoisting machine is provided in plurality, The learning data storage unit stores learning data for each of the plurality of hoists. The hoist system of claim 8.
11. A state estimation device that estimates a state of a hoisting machine that includes a motor and controls the motor to hoist and lower a load, An operation information acquisition unit that acquires operation information of the motor; a learning data storage unit that stores learning data in which the operation information acquired by the operation information acquisition unit is associated with information indicating a state related to the hoist; a learning unit that learns a state estimation equation by using learning data stored in the learning data storage unit; a learning result storage unit that stores the state estimation equation learned by the learning unit; A state estimation unit that estimates a state related to the hoisting machine by using the operation information acquired by the operation information acquisition unit and a state estimation formula stored in the learning result storage unit; Equipped with The operation information includes both information indicating an input value to the motor and information indicating an output value from the motor. State estimator.
12. A learning data setting unit is provided which, when the motor is controlled in accordance with a predetermined operating condition of the hoist in a test performed at the time of shipment of the hoist or at the time of inspection of the hoist, associates the operation information of the motor acquired by the operation information acquisition unit with the predetermined operating condition and sets the information as learning data; The learning data storage unit stores the learning data set by the learning data setting unit. The state estimation device according to claim 11.
13. The predetermined operating conditions include at least one of a maximum load condition of the hoist, a rated load condition of the hoist, and a no-load condition in which no load is attached to the hoist. The state estimation device according to claim 12.
Citation Information
Patent Citations
Safety monitoring system and detecting method based on modular hoisting trolley
CN110028004A
Berth crane automating device
JP2000191279A
Electrically driven hoisting machine with dynamic lift off stopping mechanism
JP2012066893A
Crane
JP2020152529A
Wire rope abnormality detection system, method, winch, and crane
JP2022144522A
