RU2783937C1 - Method, device and system for production facilities control - Google Patents

Abstract

FIELD: production facilities management.

SUBSTANCE: invention relates to the management of production facilities using real-time simulation. In particular, a method is claimed for controlling a production facility that includes a plurality of systems. The method comprises the steps of: generating, in a modeling block, a system model for one or more systems from a plurality of production facility systems based on an input signal, sensor data, and an output signal from a plurality of systems at the production facility; forming an object model for the production object based on the dependencies between the specified system models containing connections for feedback between the system models; control the production object by simulating the operation of the production object using the object model. At the same time, the control of the production object contains at least one of the stages, at which: predicting optimized input signals for the production object based on anomalies in the operation of the production object and associated state parameters that cause anomalies, while anomalies are identified based on sensor data and output signals of the object producing using one or more predetermined thresholds or machine learning technologies, and wherein the associated state parameters are determined by modeling likely abnormal conditions using a plant model using root cause technologies; validating the optimized inputs by generating simulated instances for the optimized inputs for the production target using the target model; predicting the operation of a new system to be deployed at the production facility, wherein the new system includes a newly commissioned system or a system, from a plurality of systems, with one of a hardware upgrade and a software upgrade, and initiating operation of the production facility based on the validated optimized input signals.

EFFECT: improving the efficiency of production facilities management with the possibility of remote monitoring and real-time modeling.

13 cl, 10 dwg

Description

State of the art

The present invention relates to the management of production objects using simulation. In particular, the present invention relates to the control of production facilities (producing installations) using real-time simulation.

State of the art

In general, a mining site includes a plurality of systems for performing mining. Systems include synchronous and asynchronous electromechanical machines. The mining object may typically be located in a remote area. Additionally, mining can be dangerous. Accordingly, manual observation of a target can be dangerous and time consuming.

Accordingly, the systems at the production site can be simulated. The operation of modeling systems and synchronous and asynchronous machines provides the possibility of remote monitoring of the state of the production object.

One approach may be to use offline simulations to determine the condition of the target. Such simulation methods may require significant simulation time and therefore cannot simulate the operation of a production facility in real time. Accordingly, real-time decisions cannot be made based on off-line simulations.

The essence of the invention

Real-time condition monitoring may be possible through improved production modeling. According to a first aspect of the present invention, a method for controlling a target is disclosed. A mining facility includes a variety of systems that are used to perform mining operations. The method includes generating, in a modeling unit, system models for one or more systems from a plurality of production facility systems; wherein the system models are formed based on one of an input signal, sensor data, and an output signal from a plurality of systems at the production site; generating an object model for the production object based on the dependencies between the system models, wherein the dependencies contain connections for feedback between the system models; and controlling the production object by simulating the operation of the production object using the object model, while the management of the production object (150A-N) comprises at least one of: predicting optimized input signals for the production object (150 A-N) based on the simulation of the production object; and predicting the operation of a new system to be deployed at the production site (150 A-N), wherein the new system includes a newly commissioned system or a system from a plurality of systems, with one of a hardware upgrade and a software upgrade.

A mine may refer to any mining operation that may be used for mineral extraction, offshore mining, or metal mining. Examples of mining facilities include mining operations for minerals such as iron ore, copper ore, and so on. At the mining site, the mined rocks can be placed in large ore crushers with diameters of approximately 13-15 meters. In some cases, 400,000 tons of ore are mined annually. The mining object includes motors and drives that drive the grinding operation. An exemplary motor includes an annular motor and an exemplary crusher drive includes a gearless crusher drive rated up to 24 MW. In another example, the motor is a single/double gear drive for crushers, mine hoists and conveyors. Accordingly, a plurality of systems at a production site include a motor and a drive. Other systems include a power supply unit, a transformer unit, a converter unit, and a load unit. The load block may further include a power interruption block and a crushing load.

In an embodiment, the modeling of the production facility is performed by modeling the scheme of systems at the production facility. For example, the systems include a drive, a converter, or a conveyor line, and a systems diagram simulation is used to simulate the operation of a production facility based on dependencies between the diagram simulations of each of the systems.

Controlled converters such as cycloconverters are used to control the drive. The advantage of the cycloconverter is its high availability, high overload capacity and high efficiency. Cycloconverters are complex and therefore not easy to model. The present invention advantageously simulates the operation of a cycloconverter to determine control parameters in real time. Additionally, the present invention advantageously models a power supply unit, a transformer, a motor, a drive unit, and a load unit, and thereby forms a digital copy of a mining object.

As used herein, the term "system models" refers to virtual copies of systems at the production site. Models of the system can be implemented based on the physics that determines the system (physical model). System models can include computer-aided design (CAE) drawings, 1D or 3D models for computational fluid dynamics (CFD), noise, vibration and harshness (NVH), structures, and the like. In some embodiments, the system models may be implemented using one or more machine learning models in combination with a physical (physics-based) model.

For example, in an embodiment, the input of the transformer block and the output of the transformer block are analyzed based on the physics behind the transformer block to form a transformer model. In another embodiment, sensor data associated with the transformer block is used as feedback to update the physical model of the transformer.

The method may include generating a supply model for the supply unit, a transformer model, a state machine model for the rectifiers in the cycloconverter, a motor model for the motor, and a load model for the load unit. The load model may further include an interrupt model for the interrupt block. System models include power supply model, transformer model, motor model, load model, and interruption model.

In an embodiment, the method may include generating state machine models for the rectifiers based on the modulation factor or firing angle of one or more thyristors in the rectifiers. In a preferred embodiment, start pulses and a function of voltage and current from a transformer unit or supply unit are used to form state machine models.

The method may include determining converter states of the cycloconverter rectifiers, the converter states being formed based on trigger pulses injected into the thyristors in the rectifiers. Converter states are used to form a state machine model for rectifiers. In order to determine the state of the transducer, the state values may need to be predicted. Accordingly, the method may include predicting state values for the states of the converter based on the start pulse, the voltage of the thyristors, and the current of the thyristors. The method may further include multiplexing state values to determine transducer states. Additionally, the method may include predicting a state transition for the cycloconverter based on the current state of the cycloconverter.

In one embodiment, the method may include determining an equivalent circuit for the transducer states. For example, a thyristor can be modeled in terms of voltage, current, and resistance. Additionally, various thyristors can be "on" or "off". Accordingly, an equivalent circuit for each converter state can be formed. One skilled in the art will appreciate that the thyristors do not turn off until the current is zero. Therefore, the formation of a finite state machine model is complex. An advantage of the present invention is that the states of the converter are predicted based on the voltage measured at the ends of the thyristors. Additionally, in an embodiment, conditions such as erroneous pulses and/or double commutations may be modeled based on historical data (history data, historical data) associated with the cycloconverter.

To ensure that the complexities of the systems are modeled accurately, the method may include validating the system models by collaboratively modeling the system models on one or more simulation platforms.

For example, a system model may be generated on a circuit simulation platform and validated on a custom block-based dynamic simulation platform. To ensure that the co-simulation is correctly performed, the method may include synchronizing the co-simulation of the system models to the simulation platforms.

In another embodiment, validation of system models may be performed by comparing simulated outputs and outputs from systems at the production site. The difference in the simulated outputs and outputs is used as feedback to update the system model.

The system models depend on each other. The dependencies are modeled to form an object model for the production object. Accordingly, as used herein, an "object model" refers to a virtual copy of a mining object. The object model provides an opportunity to observe the state of the production object. Additionally, the object model is mainly used to predict and analyze anomalies at the production object.

The method may include determining dependencies between system models based on at least one of engineering drawings, process flow diagrams, a layout map of a production site, and relationships between data points in sensor data. For example, the process and instrumentation schemes of a production site can be used to determine systems that are associated with each other. In another example, sensitivity analysis may be performed on sensor data to determine the relationship between data points. The relationship between data points can be used to define associations between systems.

In an embodiment, dependencies are determined such that a state machine model for the rectifiers in the cycloconverter is generated based on the motor current output from the motor model provided as input to the state machine model. In another embodiment, the engine model for the engine is generated based on the speed output from the load model provided as input to the engine model.

The formation of real-time system models may be required to monitor the operation of the production facility without any delay. Accordingly, the method may include generating one of a state machine model, a motor model, and a load model using a field programmable gate array (FPGA).

The object model can be used for remote monitoring and control of the anomaly. The method may include identifying anomalies in the operation of the production object based on sensor data and output signals. Anomalies can be identified using a predetermined threshold value for sensor data and output signals. In another embodiment, anomalies may be identified using machine learning algorithms such as k-means clustering, range plots, and so on.

The method may further include determining the state parameters causing the anomalies by modeling possible anomaly conditions with the object model. To model possible anomaly conditions, the method may include predicting possible anomaly conditions based on input data, sensor data, and output data using root cause analysis algorithms such as Bayesian networks.

The method may include predicting optimized inputs for a target based on anomalies and state parameters. In an embodiment, the state parameters and the optimized input signals are compared in a logarithmic table.

In certain scenarios, optimized inputs are predicted based on trial and error. Accordingly, it is useful to validate the optimized inputs and impact on the target before updating. The method may include validating the optimized inputs by generating simulation instances for the optimized inputs for the production target using the target model. For example, simulation examples are generated when optimized input signals are provided as input to an object model. The behavior of the object model is analyzed by determining whether the output values of the system models fall within a threshold. The method may further include operating the production target based on the validated optimized inputs.

In another scenario, there may be a hardware or software upgrade of the target. This update is called a "new system". For example, hardware upgrades include modifications to the physical parts of the type, such as replacing a thyristor in a cycloconverter or replacing motor windings for a motor. Software updates include modification of cycloconverter control parameters, update in control software, etc.

The method may further include predicting the performance of a new system to be deployed at a production site using the site model. As mentioned above, a new system includes a newly commissioned system or a system from a plurality of systems, with one of a hardware upgrade and a software upgrade. The method may further include optimizing the design parameters of the new system based on the predicted performance. For example, the new system may refer to a new modulation method for triggering thyristors. The firing angle (starting angle) can be optimized to result in a reduction in reactive power.

The method may include simulating the commissioning steps on a plant model. Modeling of the stages of commercial production can be used to compare the operation of the production facility against existing guidelines.

The method may further include modeling the rate control parameters from the plant model to experience a plurality of rate control modes. The method may also include displaying anomalies associated with rate control modes. Additionally, the method includes identifying changes in rate control parameters to mitigate anomalies.

The second aspect of the present invention is a simulation unit for controlling a production target. The simulation block contains a field programmable gate array (FPGA). A memory communicatively coupled to the FPGA, the memory comprising a simulation module stored in the form of machine readable instructions executable by the FPGA, the simulation module being configured to perform one or more of the method steps described above. The simulation module includes system models and an object model. In an embodiment, the simulator may include a combination of an FPGA and a CPU to execute the object model.

In an embodiment, the simulator may be located remotely from the production site and connected with the possibility of communication through the network. For example, sensor data from a production site is received by the simulator via a network. The simulation block executes a model of the object using sensor data to determine anomalies or new design parameters for the production object.

A third aspect of the present invention is a system for managing at least one production target. The system comprises one or more devices configured to provide sensor data and output data associated with the operation of at least one production target, and one or more of the aforementioned simulators connected with the possibility of communication with one or more devices. The simulation blocks are configured to manage at least one production target.

In an embodiment, the system may include a cloud computing platform communicatively coupled to one or more devices. Simulation modules of simulation blocks can be stored on a cloud computing platform and accessed through a network interface. The network interface may be a combination of a wired and a wireless data connection such as, for example, a WLAN, 2G, 3G, 4G or 5G network. Therefore, the system models and the object model can be accessed through the network interface.

A fourth aspect of the present invention is a computer-readable medium having computer-readable instructions stored thereon which, when executed by a processor, instruct the processor to perform the above steps of the method.

The present invention is not limited to a single computer system platform, processor, operating system, or network. One or more aspects of the present invention may be distributed among one or more computer systems, such as servers, configured to provide one or more services to one or more client computers, or to perform a complete task on a distributed system. For example, one or more aspects of the present invention may be implemented on a client/server system that includes components distributed among one or more server systems that perform a variety of functions in accordance with various embodiments. These components contain, for example, executable, intermediate or interpreted code that is transmitted over the network using a communication protocol.

The above and other features of the invention will not be discussed with reference to the accompanying drawings of the present invention. The illustrated embodiments are intended to illustrate, but not to limit the invention.

The present invention is further described hereinafter with reference to the illustrated embodiments shown in the accompanying drawings, in which:

Fig. 1 illustrates a block diagram of a system for managing production targets, according to an embodiment of the present invention;

Fig. 2-4 illustrate the formation of a state machine model for a rectifier in a cycloconverter, according to an embodiment of the present invention.

Fig. 5 illustrates a block diagram of the facility model for the production facility of FIG. 1 according to an embodiment of the present invention;

Fig. 6 illustrates a block diagram of a simulation block for controlling a production object, according to an embodiment of the present invention;

Fig. 7 illustrates a method for controlling a target, according to an embodiment of the present invention; and

Fig. 8, 9 and 10 are graphs that illustrate the optimization of input signals using the plant model, according to an embodiment of the present invention.

Hereinafter, embodiments for carrying out the present invention are described in detail. Various embodiments are described with reference to the drawings, with like reference numerals being used to refer to like elements throughout the drawings. In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of one or more embodiments. It may be obvious that such embodiments may be practiced without these specific details.

Fig. 1 illustrates a block diagram of a system 100 including a simulator 102 for controlling production targets 150A-150N, in accordance with an embodiment of the present invention. Production objects 150A-150N include devices 152A-152N that generate and transmit sensor data associated with the respective production object to modeling unit 102. In an embodiment, devices 152A-N are part of system 100.

Devices 152A-152N measure the operating parameters of the respective mining objects 150A-150N. The term "operating parameter" refers to one or more characteristics of the target. For example, operating parameters include ambient temperature, ambient pressure, vibration data from site systems, voltage, magnetic flux, and so on.

Production facilities 150A-150N also include a variety of systems for performing various mining operations. The systems include a supply unit 154A-154N, a transformer unit 156A-156N, a cycloconverter 158A-158N, an annular motor 160A-160N, and a load unit 162A-162N. The power supply unit 154A-154N may also be referred to as a power supply network 154A-154N and may include a medium voltage switchgear. The load block 162A-1060N includes the braking systems and crusher load of the respective mining object.

System 100 includes a simulation unit 102 that communicatively communicates with production targets 150A-150N via a network interface 140. Network interface 140 may include a wired data connection, such as Ethernet, and/or a wireless data connection. , such as, for example, a WLAN, 2G, 3G, 4G or 5G network.

The simulation block 102 includes a transceiver 104 and one or more computing blocks 106 implemented as field programmable gate arrays (FPGAs). In an embodiment, computing units 106 may include a combination of multiple FPGAs and central processing units (CPUs). Simulator 102 also includes a graphical user interface (GUI) 108 to enable a user to control simulation 102. Simulation unit 102 also includes a memory 110 configured to store computer program instructions defined by modules, such as simulation module 125 .

Network interface 140 also provides access to cloud computing platform 130 for communication with production facilities 150A-150N. As used herein, the term "cloud computing environment" refers to a processing environment containing configurable computing physical and logical resources, such as networks, servers, storage, applications, services, etc., and data distributed over a network 140. Cloud computing platform 130 is configured with memory and a processor to store and execute simulation module 125 .

Modeling module 125, when executed, is configured to generate object models for production objects 150A-150N. Simulation module 125 includes modules 112, 114, 116, 118, and 120 to generate system models for systems 154A-N, 156A-N, 158A-N, 160A-N, and 162A-N at production sites 150A-150N.

The simulation module includes a power supply model generator 112 to generate power supply models for power supply units 154A-N. The power supply models are formed based on at least the source voltage and the power supply voltage (ie, 11, 20, 30 kV). Additionally, power supply models may also be generated based on predetermined current/voltage relationships measured at nodes in power supply units 154A-N.

Modeling module 125 includes a transformer model generator 114 to generate a transformer model for transformer blocks 156A-N. The transformer model is generated based on the power supply voltage measured at the output of the power supply units 154A-N and the transformer voltage measured at the output of the transformer units 156A-N. In an embodiment, the transformer model is generated based on the electrical behavior of the internal components of the transformer unit.

Simulation module 125 also includes a converter model generator 116 configured to generate state machine models for the rectifiers in cycloconverter 158A-N. In some embodiments, production targets 150A-N may include drive rectifiers. In such embodiments, converter model generator 116 is configured to generate drive models for the drive rectifiers. In an embodiment, the converter model generator 116 runs on multiple FPGAs to enable fast real-time simulation of the cycloconverters 158A-N.

The finite state machine models are formed based on at least the firing pulses of one or more thyristors of the 158A-N cycloconverters. Additionally, state machine models are also generated based on a function of voltage and current from either the transformer units 156A-N or the power supply unit 154A-N. The generation of the state machine model is further explained in FIG. 1B, 1C and 1D.

Simulation module 125 includes an engine model generator 118 to generate engine models for engine 160A-N based on the speed output from the load models.

Simulation module 125 further includes a load model generator 120 . Load model generator 120 is configured to generate load models for load blocks 162A-N. Load model generator 120 is further configured to generate a brake model for braking systems. The load models are generated based on the torque output from the 160A-N motor. Modeling module 125 includes a validation module 122 . Validation module 122 is configured to validate power supply models, transformer models, state machine models, and load models (collectively referred to as system models). Validator 122 is configured to jointly model the system models to ensure that the system models act as virtual copies of the associated system.

Object models are formed from system models. Modeling module 125 is configured to determine dependencies between system models. Dependencies are determined by means of physical relationships between system models. Dependencies between system models are determined based on at least one of the engineering drawings, process flow diagrams, production layout map, and relationships between data points in sensor data. Fig. 5 illustrates the formation of object models based on dependencies between system models.

The object models may be displayed on the GUI 108 via graphical representations. Production objects 150A-150N are controlled by analyzing graphical representations. In some embodiments, mining objects 150A-150N may be managed by detecting anomalies in the operation of mining objects 150A-150N.

In some embodiments, memory 110 and cloud computing platform 130 may further include a prediction module 126 and an optimization module 128. Prediction module 126 is configured to predict the performance of new systems that may be added to production targets 150A-N. Additionally, optimization module 128 is configured to predict and validate optimized inputs to improve performance of production facilities 150A-N.

The GUI 108 is used to display the predicted performance of new systems. Additionally, the GUI 108 is used to display the operation of the target in order to validate the optimized inputs. In an embodiment, the modeling unit 102 may be communicatively coupled to a display device configured to display 3D interactive models of production targets 150A-N simultaneously or sequentially.

Fig. 2-4 illustrate the formation of a state machine model for rectifiers in a cycloconverter 158A, according to an embodiment of the present invention.

Fig. 2 illustrates a circuit diagram 170 of a rectifier in a cycloconverter 158A. The circuit diagram illustrates a thyristor converter bridge 172 with thyristors 1, 2, 3, 4, 5 and 6. The supply voltage is a three-phase power supply with voltage U1, U2, U3, inductance Lk and current I1, I2 and I3. The current Id and the output voltage UA are observed at the ends of the converter bridge 172 on the thyristors.

Fig. 3 illustrates a state transition matrix 180 with active thyristors 182 and associated converter states 184. Matrix 180 illustrates the transition from one active thyristor to another. The thyristor is activated by starting pulses or by changing the currents I1, I2 and I3. Additionally, one of ordinary skill in the art would appreciate that states 184 may change based on voltage phases. Additionally, the states 184 may change if the thyristor turn-on is not normal. In addition, in the case of erroneous pulses and double switching, states 184 may change. Accordingly, state change matrix 180 is an embodiment that can be updated to include the above changes.

Fig. 4 is an equivalent circuit 190 for the rectifier in the cycloconverter 158A in state 8 (indicated by reference numeral 186). In state 8, thyristor 1 is active and current flows from thyristor 2 to thyristor 6. Thyristors 1, 2, and 6 are modeled with a deflection characteristic. Additionally, as illustrated, thyristors 1, 2 and 6 are modeled as a combination of resistance (RD) * current and voltage U _FOD . Additionally, the inductance Lk is modeled as the inductance and resistance Rk.

Equivalent circuit 190 may be used to determine the value of the output voltage UA. For example, the following equation is used to determine the output voltage UA.

The voltage at the ends of the inductance Lk and resistance Rk is determined as follows for the three-phase voltage U1, U2, U3.

Additionally, the thyristor voltage at the ends of thyristors 1, 2, 3, 4, 5 and 6 is determined as follows.

The above determination can be made for all states 184. In an embodiment, the above determination is made in parallel for all states 184, and the output voltage UA for the rectifier is determined by selection with a multiplexer.

Fig. 5 illustrates a block diagram of a facility model 200 for production facility 150A in FIG. 1 according to an embodiment of the present invention. Object model 200 includes system models 210, 220, 230, and 240. For the purpose of FIG. The 5 system models include a transformer model 210, a state machine model 220, a motor model 230, and a load model 240. Additionally, as shown in FIG. 5, dependencies between system models are illustrated using feedback connections 238 and 244.

The transformer model 210 is formed based on the power supply voltage 212 measured at the output of the power supply unit 154 . The transformer voltage 214 is output from the transformer model 210 and is comparable to the voltage measured at the output of the transformer block 156A. The transformer model 210 is formed based on the electrical behavior of the internal components of the transformer unit 156A.

State machine model 220 is generated based on start pulses 216, transformer voltage 214, and motor current 238 from motor model 230 using circuit diagram 190 and state transition matrix 180.

Additionally, motor model 230 is generated based on output voltage (UA) 226 from state machine model 220 and speed 244 derived from load model 240. Load model 240 receives torque 236 output from engine model 230 to generate speed 244.

One of ordinary skill in the art will appreciate that the relationship in FIG. 5 is an example. A plurality of such dependencies may be generated from engineering drawings, process flow diagrams, layout maps of production facility 150A, and so on. In an embodiment, sensor data from devices 152A is analyzed by performing sensitivity analysis to determine dependencies in system models.

Fig. 6 illustrates a block diagram of a simulator 330 for controlling a production facility 310, according to an embodiment of the present invention. Similar to the production targets in Fig. 1, the production facility 310 includes a power supply unit, a transformer unit, a cycloconverter control unit 312, a motor, and a load unit. Simulator 330 may be communicatively connected to production target 310 via a network interface.

The cycloconverter control unit 312 is configured with a power rating of 3.2-40 MVA. The output voltage can vary from 1500-4000 V. Additionally, the thyristors in the cycloconverter 312 are placed as three-phase bridges in a 6-pulse, 12-pulse, or 24-pulse connection.

The cycloconverter 312 can be controlled and commissioned by various devices and computer program instructions. In an embodiment, production facility 310 also includes a graphical user interface (GUI) 314, an automation module 316, and a commissioning module 318.

The GUI 314 is used to display and visualize the operation of the cycloconverter 312. For example, thyristor triggers, input voltage, output voltage, etc. can be displayed. The automation module 316 is configured to display interconnections at the production facility 310 . Automation module 316 may also be used to indicate to an operator if an anomaly exists at a production site. Commissioning module 318 is generally used during commissioning of cycloconverter 312. Commissioning module 318 may also be used to implement diagnostics via GUI 314.

In addition, the production object 310 further includes a hardware interface 320 allowing communication between the simulator 330 and the cycloconverter 312.

Simulator 330 includes signal preconditioner 332, processing blocks 334 and 336, and GUI 338. Signal preconditioner 332 may include I/O boards for analog and digital signal. To enable fast simulation, the signal preconditioner 332 includes dedicated boards. Signal preconditioner 332 is configured to provide signal adaptations to specific electrical system voltages (12V, 24V, etc.); also for current interfaces.

Processing units 334 and 336 include a field programmable gate array (FPGA) 334 and central processing units (CPUs) 336. FPGA 334 can be used for fast online simulation of the systems at the production site 310. Additionally, the CPU 336 can be used for offline simulation. For example, the state machine model for the rectifiers of the cycloconverter 312 may be executed on the FPGA 334. The load model for the load block may be executed on the CPU 336. The use of processing blocks 334 and 336 is performed based on a requirement for fast online simulation.

GUI 338 is configured to display output from processing blocks 334 and 336 . Additionally, the GUI 338 is used to change the parameters associated with the models implemented on the blocks 334 and 336 processing.

Fig. 7 illustrates a method 400 for controlling a target, in accordance with an embodiment of the present invention. Method 400 begins at 402 with receiving input, sensor data, and output regarding a plurality of systems at a production site.

At step 404, a power supply model for the power supply unit and a transformer model for the transformer unit are generated. In an embodiment, step 404 may be performed offline prior to commissioning the power supply unit and the transformer unit. Additionally, the power supply model and the transformer model can be updated after installation and during production operation.

At 406, a state machine model for the cycloconverter and a motor model for the motor are generated. In an embodiment, step 406 is performed in the FPGA to enable fast real-time simulation of the cycloconverter and motor in-loop hardware. As used herein, the term "in-loop hardware simulation" refers to a method step in which one or more real components interact with components that are simulated in real time. Accordingly, the state machine model receives input from a cycloconverter deployed at the production site. The output of the state machine model can be further used to modify the inputs to the cycloconverter.

In an embodiment, step 406 may include generating a state machine model for the cycloconverter rectifiers based at least on start pulses for one or more rectifier thyristors and a function of voltage and current from at least the transformer unit and the power supply unit.

Step 406 may include determining converter states for the thyristors of the rectifiers in the cycloconverter, the converter states being generated based on trigger pulses input to the cycloconverter. Transducer states are used to form a finite state machine model. In order to determine the state of the transducer, the state values may need to be predicted. Accordingly, step 406 may include predicting the state values for the states of the converter based on the start pulse, the voltage of the thyristors, and the current of the thyristors.

Block 406 may further include multiplexing state values to determine transducer states. Additionally, step 406 may include state change prediction for the rectifiers based on the current state of the cycloconverter rectifiers.

In an embodiment, step 406 may include determining an equivalent circuit for the transducer states. For example, a thyristor can be modeled in terms of voltage, current, and resistance. Additionally, various thyristors can be "on" or "off". Accordingly, an equivalent circuit for each converter state can be formed.

Additionally, in an embodiment, conditions such as erroneous pulses and/or double switches may be modeled based on historical data associated with the cycloconverter. In such an embodiment, the simulation may be performed offline by the CPU.

At step 408, a load model and a power interruption model for the load block and the chopper block are generated. In an embodiment, step 408 is performed on the CPU.

The models generated in blocks 404, 406, and 408 are collectively referred to as system models.

At block 410, the system models are validated by co-simulating the system models on one or more simulation platforms. For example, a system model may be generated on a circuit simulation platform and validated on a custom block-based dynamic simulation platform. To ensure that the co-simulation is correctly performed, step 410 may include synchronizing the co-simulation of the system models to the simulation platforms.

In another embodiment, step 410 may be performed by comparing simulated outputs from system models and outputs from systems at the production site. The difference in the simulated outputs and outputs is used as feedback to update the system models.

The formation of real-time system models may be required to monitor the operation of the production facility without any delay. Accordingly, step 410 may include generating one of a state machine model, a motor model, and a load model using a field programmable gate array (FPGA).

At block 420, the dependencies between the system models are used to generate an object model for the production object. Accordingly, as used herein, the term "object model" refers to a virtual copy of a mining object. The object model provides an opportunity to observe the state of the production object. Additionally, the object model is mainly used to predict and analyze anomalies at the production object.

Step 420 may include determining dependencies between system models based on at least one of the engineering drawings, process flow diagrams, production layout map, and relationships between data points in sensor data. For example, the process and instrumentation schemes of a production site can be used to determine systems that are associated with each other. In another example, sensitivity analysis may be performed on sensor data to determine the relationship between data points. The relationship between data points can be used to define associations between systems.

In an embodiment, dependencies are determined such that a state machine model for the cycloconverter rectifiers is generated based on the motor current output from the motor model provided as input to the state machine model. In another embodiment, the engine model for the engine is generated based on the speed output from the load model provided as input to the engine model.

At block 430, the object model is used to manage the production object. The control of the target can be performed by monitoring the remote state and managing the anomaly.

At step 432, anomalies in the operation of the production object are identified based on sensor data and output signals. Anomalies can be identified using a predetermined threshold value for sensor data and output signals. In another embodiment, anomalies can be identified using machine learning algorithms such as k-means clustering, range plots, and so on.

Step 432 may further include determining the state parameters causing the anomalies by modeling possible anomaly conditions with the object model. Step 432 may include predicting optimized inputs for the target based on anomalies and state parameters. In an embodiment, the state parameters and the optimized input signals are compared in a logarithmic table.

Step 432 may include validating the optimized inputs by generating simulation examples for the optimized inputs for the production target using the target model. For example, simulation examples are generated when optimized input signals are provided as input to an object model. The behavior of the object model is analyzed by determining whether the output values of the system models fall within a threshold. Step 432 may further include initiating operation of the production facility based on the validated optimized inputs. Additionally, the production object operates according to optimized input signals.

At step 434, the operation of the new system is predicted using the object model. The new system may include a newly commissioned system or an existing system with hardware and software upgrades. Step 434 may further include optimizing the design parameters of the new system based on the predicted performance. For example, a new system may refer to a new modulation method for triggering thyristors. The firing angle can be optimized to lead to a reduction in reactive power.

At block 436, the commissioning steps are modeled on the facility model. Modeling of the stages of commercial production can be used to compare the operation of the production facility against existing guidelines.

The implementation of steps 432, 434 and 436 can be visualized through graphs displayed on the GUI. Fig. 8-10 illustrate the implementation of steps 432 and 434.

Fig. 8, 9, and 10 are graphs that illustrate the optimization of input signals with a target model for a production target. The plant model simulation examples are used to identify the most optimal inputs and are referred to as "optimized inputs".

Fig. 8 illustrates the trend 500 of an object using graphs 502-510.

Fig. 9 illustrates a simulated trend 520 using plots 522-530.

Fig. 10 illustrate an optimized trend 540 with plots 542-550.

Plots 502, 522 and 542 are speed plots. In FIG. 8, a speed graph 502 includes a reference speed 502a and a measured speed 502b. In FIG. 9, speed plot 522 includes a simulated speed reference 522a and a simulated speed 522b. In FIG. 10, velocity plot 542 includes a reference velocity 542a for optimized inputs and a measured velocity 542b for optimized inputs.

Plots 504, 524 and 544 are drive current (EC) plots. In FIG. 8, EC plot 504 includes a reference EC 504a and a measured EC 504b. In FIG. 9, rate plot 524 includes a simulated reference EC 524a and a simulated EC 552b. In FIG. 10, EC plot 544 includes a reference EC 544a for optimized inputs and a measured EC 544b for optimized inputs.

Plots 506, 526, 546 are magnetizing current plots. Plots 508, 528 and 548 are torque current plots. Additionally, plots 510, 530, and 550 are three-phase current plots. Plots 506, 508 and 510 are measured in situ before using the optimized input signals. Plots 526, 528, and 530 are simulated plots generated from an object model. Plots 546, 548 and 550 are measured at the production site after the optimized inputs are realized.

In FIG. 8 and 5B, crusher oscillation is evident in plots 502, 508 and 522, 528, respectively. Vibrations 502 and 508 crushers are observed at the site of the mining object. The object model also indicates vibrations 522 and 528 of the crusher.

Crusher vibrations 522 and 528 were used to further analyze the state parameters associated with crusher vibrations 522 and 528. Probable causes of crusher vibrations are determined based on the engineering drawings of the mining site.

For example, in the present case, the technical drawings of the mining site indicate a crusher drive without gearing, which is held by brakes in an unbalanced position. Accordingly, the probable cause for the oscillator 522 and 528 may be related to the operation of the brakes. The object model is used to simulate anomalies in the operation of the brakes. In the present example, it is determined that a delay in opening the brakes can cause the crusher to vibrate. Accordingly, the state parameters for the brake release time have been identified. Additionally, the delayed release of the brake is modeled on the target and compared to plots 522 and 528.

When state parameters are identified, optimized input signals are predicted based on the anomaly and state parameters. For the delayed brake release example, the speed controller values are identified as inputs to be optimized. Additionally, multiple speed controller values are modeled from the plant model. The implementation of the optimized input signals is illustrated in FIG. 10, speed plot 542 and torque plot 548 indicate reduced crusher oscillation at 552 and 554.

In operation, the optimized inputs to the speed controller are passed to the commissioning engineer at the production site. The commissioning engineer implements the optimized input signals. In another embodiment, the optimized inputs are implemented using a cloud computing platform where the relevant variables are updated. The updated variables are reflected in the production target. The present invention may take the form of a computer program product comprising program modules accessible from a computer-usable or computer-readable medium storing program code for use by or in connection with one or more computers, processors, or instruction execution system. For the purposes of this description, a computer-usable or computer-readable medium can be any device that can contain, store, transmit, distribute, or carry a program for use by or in connection with an instruction execution system, equipment, or device. The medium may be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system (or equipment or device) or propagation media per se, since signal carriers are not included in the definition of physical computer-readable media, may include semiconductor or solid-state memory, a magnetic tape, a removable computer floppy disk, a random access memory (RAM), a read only memory (ROM), a hard magnetic disk, and an optical disk such as a compact disk read only memory (CD-ROM), a readable/writable compact disk, and DVD. Both the processors and the software code for implementing each aspect of the technology may be centralized or distributed (or a combination thereof), as is known to those skilled in the art.

While the present invention has been described in detail with reference to certain embodiments, it should be understood that the present invention is not limited to these embodiments. In view of the present disclosure, many modifications and variations will be presented to specialists in the field of technology without departing from the scope of various embodiments of the present invention, as described in this document. The scope of the present invention is therefore defined by the following claims and not by the preceding description. All changes, modifications and variations falling within the meaning and range of equivalence of the claims are to be considered within the scope of the claims. All preferred embodiments claimed in method claims may also apply to system/device claims.

Claims (38)

1. A computer-implemented method for controlling a production facility (150 A-N), including a plurality of systems, the method comprises the steps of: forming, in the modeling block (102, 330), system models (210, 220, 230, 240) for one or more systems from a plurality of production object systems (150 A-N); wherein the system models (210, 220, 230, 240) are formed based on an input signal, sensor data, and an output signal from a plurality of systems at the production site (150 A-N); an object model (200) is formed for the production object (150 A-N) based on the dependencies between the models (210, 220, 230, 240) of the system, while the dependencies contain connections (228, 238, 244) for feedback between the models (210, 220 , 230, 240) systems; control the production object (150 A-N) by simulating the operation of the production object (150 A-N) using the object model (200), while the management of the production object (150 A-N) contains at least one of the stages, in which: predicting optimized input signals for the production object (150 A-N) based on anomalies in the operation of the production object and associated state parameters that cause anomalies, while anomalies are identified based on sensor data and output signals of the production object using one or more predefined threshold values or technologies machine learning, wherein the associated state parameters are determined by modeling likely abnormal conditions using an object model using root cause technologies; validate the optimized input signals by generating simulated instances for the optimized input signals for the production object (150 A-N) using the object model (200); predicting the operation of a new system to be deployed at the production site (150 A-N), wherein the new system includes a newly commissioned system or a system from a plurality of systems, with one of a hardware upgrade and a software upgrade, and initiating operation of the production facility (150 A-N) based on the validated optimized input signals. 2. The method according to claim 1, wherein one or more systems include a power supply unit (154A-N), a transformer unit (156A-N), a cycloconverter (158A-N), a motor (160A-N), and a unit ( 162A-N) loads, and at the same time, the formation of the system model includes the steps at which: forming a model (220) of a state machine for cycloconverter rectifiers (158A-N) based on at least start pulses (216) of at least one cycloconverter thyristor (158A-N) and a function of voltage (214) and current (238), at least from the block (156A-N) of the transformer and model (230) of the engine; generating a motor model (230) for the motor (160A-N) as a function of at least voltage (226) of the cycloconverter (158A-N), rotation speed (244), motor current (238), and motor torque (236); and a load model (240) is formed for the load block (162A-N) based on a function of at least the engine torque (236) and rotation speed (244), while the system models (210, 220, 230, 240) are included in itself a state machine model (220), a motor model (230), and a load model (240). 3. The method according to claim 2, wherein the generation of the FSM model (220) comprises the step of determine the states of the converter for the thyristors of the rectifiers, the states of the converter are formed on the basis of starting pulses input to the cycloconverter (158A-N). 4. The method according to claim 3, wherein the formation of the states of the converter comprises the steps, in which: predicting state values for the converter states based on the start pulse, the thyristor voltage, and the thyristor current; and determining the transducer states based on the multiplexed output value for the state values for the transducer states. 5. The method according to any one of paragraphs. 2-4, further comprising the step of predicting a state change for the cycloconverter (158A-N) based on the current state of the cycloconverter (158A-N). 6. A method according to any one of the preceding claims, further comprising the step of validating the system models (210, 220, 230, 240) by co-simulating the system models (210, 220, 230, 240) on one or more simulation platforms. 7. The method according to claim 1, wherein the formation of a model (200) of an object for a production object (150 A-N) comprises the step of: dependencies between system models (210, 220, 230, 240) are determined based on at least one of the engineering drawings, process flow diagrams, production facility (150 A-N) layout map, and relationships between data points in sensor data. 8. The method according to any of the preceding paragraphs, at least according to paragraphs. 2 and 7, further comprising the steps of: generating a state machine model (220) for the cycloconverter rectifiers (158A-N) based on the motor current output from the motor model (230) provided as input to the state machine model (220); and generating an engine model (230) for the engine (160A-N) based on the speed derived from the load model (240) provided as input to the engine model (230). 9. A method according to any one of the preceding claims, further comprising the steps of one of the state machine model (220), motor model (230), and load model (240) is formed using a field programmable gate array (FPGA). 10. A method according to any one of the preceding claims, further comprising the steps of: predicting the operation of a new system to be deployed at the production facility (150 A-N) using the facility model (200); and optimize the design parameters of the new system based on the predicted performance. 11. Block (102, 330) simulation to control the object (150 A-N) production, block (102, 330) simulation contains: user-programmable gate array (FPGA); and a memory communicatively coupled to the FPGA, the memory comprising a simulation module stored in the form of machine-readable instructions executable by the FPGA, the simulation module being configured to perform one or more of the steps of the method of claims. 1-10. 12. System (100) for managing at least one production facility (150 A-N), the system contains: one or more devices configured to provide sensor data and output data associated with the operation of at least one object (150 A-N) production; and one or more simulators (102, 330) of claim 11 communicatively connected to one or more devices, wherein the simulators (102, 330) are configured to control at least one production target (150 A-N). 13. A computer-readable medium having machine-readable instructions stored thereon, which, when executed by the processor, cause the processor to perform the method of claims. 1-10.

Images