Device parameter adjustment method, device, electronic device and storage medium
Through the equipment parameter adjustment method based on artificial intelligence, the equipment parameters are automatically adjusted using pre-trained models, and the problems of low efficiency and low accuracy caused by relying on professional experience in the existing technology are solved, and efficient dynamic adjustment of equipment parameters is achieved. Reduced costs.
Patent Information
- Application Number
- CN202411125038.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-16
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-08-16
AI Technical Summary
In the prior art, equipment parameter adjustment is highly dependent on the experience of professionals, resulting in low efficiency, low accuracy, and high time and labor costs.
The device parameter adjustment method based on artificial intelligence is adopted, and the current device status and the expected device status are obtained, and the parameter adjustment strategy model is used to adjust the device parameter to achieve automatic adjustment.
It improves the efficiency and accuracy of equipment parameter adjustment, reduces time and labor costs, and realizes efficient and dynamic adjustment of equipment parameters.
Smart Images

Figure CN118643621B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to a device parameter adjustment method, device, electronic device and storage medium. Background Art
[0002] In the field of industrial manufacturing, industrial equipment plays a key role in ensuring the smooth operation of enterprise production. The key to the successful operation of equipment is to ensure that they meet pre-defined performance indicators. However, due to the high complexity of industrial equipment systems, there are a large number of parameters that need to be adjusted, and there are complex nonlinear and dynamic relationships between these parameters, so constant adjustments are required to ensure that the equipment can operate efficiently. In addition, in order to achieve optimal performance, the equipment needs to be dynamically adjusted in a timely manner according to changes in external conditions.
[0003] In the process of realizing the present invention, the inventors found that the prior art has the following defects: At present, the adjustment of equipment parameters is highly dependent on the experience of professionals, which not only consumes a lot of manpower and time, but also human uncertainty may lead to the decline of equipment performance and uncertainty of product quality. Considering that a factory may contain a variety of equipment, the problem is more complicated, resulting in low efficiency and accuracy of equipment parameter optimization, and high time and labor costs. Summary of the invention
[0004] The present invention provides a device parameter adjustment method, an apparatus, an electronic device and a storage medium to improve the efficiency and accuracy of device parameter adjustment and reduce time cost and manpower cost.
[0005] According to one aspect of the present invention, a method for adjusting device parameters is provided, which includes:
[0006] Acquire a current device state of a target device and a device description parameter to be adjusted corresponding to the current device state; wherein the target device is an artificial intelligence-driven device;
[0007] Get the expected device state that the target device expects to obtain;
[0008] The current device state and the expected device state are jointly input into a pre-trained device parameter adjustment strategy model to adjust the device description parameters to be adjusted, thereby obtaining the adjusted device description parameters;
[0009] The adjusted device description parameters are fed back to complete the adjustment process of the device description parameters to be adjusted.
[0010] According to another aspect of the present invention, there is provided a device parameter adjustment apparatus, comprising:
[0011] A current device state and device description parameter acquisition module, used to acquire the current device state of a target device and the device description parameters to be adjusted corresponding to the current device state; wherein the target device is an artificial intelligence-driven device;
[0012] An expected device state acquisition module is used to acquire an expected device state that a target device expects to obtain;
[0013] An adjusted device description parameter determination module is used to input the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the current device description parameters to be adjusted, thereby obtaining adjusted device description parameters;
[0014] The adjusted device description parameter feedback module is used to feed back the adjusted device description parameters to complete the adjustment process of the device description parameters to be adjusted.
[0015] According to another aspect of the present invention, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the device parameter adjustment method described in any embodiment of the present invention when executing the computer program.
[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the device parameter adjustment method described in any embodiment of the present invention when executed.
[0017] The technical solution of the embodiment of the present invention obtains the current device state of the target device and the device description parameters to be adjusted corresponding to the current device state; obtains the expected device state that the target device expects to obtain; inputs the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the current device description parameters to be adjusted to obtain the adjusted device description parameters; and feeds back the adjusted device description parameters to complete the adjustment of the current device description parameters to be adjusted. The present invention solves the problem of low efficiency and low accuracy caused by the large amount of manpower, material resources and time cost required for device parameter adjustment, improves the efficiency and accuracy of device parameter adjustment, and reduces time cost and manpower cost.
[0018] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0020] Figure 1 is a flow chart of a device parameter adjustment method provided according to Embodiment 1 of the present invention;
[0021] Figure 2a is a detailed flow chart of a method for adjusting device parameters according to Embodiment 2 of the present invention;
[0022] Figure 2b is a structural schematic diagram of a beamline station in a device parameter adjustment method provided according to a second embodiment of the present invention;
[0023] Figure 3 is a structural schematic diagram of a device parameter adjustment apparatus provided according to Embodiment 3 of the present invention;
[0024] Figure 4 It is a schematic diagram of the structure of an electronic device provided according to Embodiment 4 of the present invention. DETAILED DESCRIPTION
[0025] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0026] It should be noted that the terms "target", "current", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0027] Embodiment 1
[0028] Figure 1 A flowchart of a device parameter adjustment method is provided for Embodiment 1 of the present invention. This embodiment is applicable to the case of adjusting device parameters. The method can be executed by a device parameter adjustment device, which can be implemented in the form of hardware and / or software.
[0029] Correspondingly, such as Figure 1 As shown, the method includes:
[0030] S110: Acquire the current device state of the target device and the device description parameters to be adjusted corresponding to the current device state.
[0031] Wherein, the target device is an artificial intelligence driven device.
[0032] The current device state may be used to describe the state of the target device. The device description parameters to be adjusted may be device description parameters collected from different dimensions.
[0033] It is understandable that each current device state corresponds to a set of device description parameters, and when the expected device state is determined, a new set of device description parameters can be further determined based on the expected device state. It is understandable that the new device description parameters are also the adjusted device description parameters corresponding to the device description parameters to be adjusted.
[0034] In addition, the target device is an AI-driven device, which can obtain the current device status and expected device status in real time, and adjust the device description parameters.
[0035] In this embodiment, various types of AI-driven devices can be processed for real-time parameter optimization. Since the devices are generally highly complex industrial equipment systems, involving a large number of parameters that need to be adjusted, and there are complex nonlinear and dynamic relationships between these parameters, continuous adjustment is required to ensure that the devices can operate efficiently. In addition, the core of the successful operation of the devices is to ensure that they meet the pre-defined performance indicators, that is, to ensure the performance indicators of the devices by adjusting the device parameters.
[0036] For example, the target device may be a beamline station. A beamline station is an experimental device used to study and apply various forms of radiation beams (such as laser beams, X-rays, synchrotron radiation, ion beams, and electron beams). They are usually used for basic research and applied research in the fields of physics, chemistry, biology, and materials science.
[0037] The beamline station system is mainly divided into beamline source, beamline delivery system and experimental station. The beamline source is responsible for providing various forms of radiation beams; the beamline delivery system is responsible for delivering and adjusting the beamline flow, mainly including vacuum valves, vacuum pumps, reflectors, monochromators and focusing mirrors. The experimental station is a platform for research and application using the adjusted beamline flow, including various detectors (spectrometers, mass spectrometers, etc.).
[0038] The equipment description parameters of the beamline station may include: beamline source parameters, beamline delivery system parameters and laboratory table parameters, etc.
[0039] In detail, the beam source parameters may include beam source energy and beam source current intensity, etc. Among them, the radiation source energy can be changed by adjusting the energy setting of the accelerator to change the energy of the beam line. The beam source current intensity can be adjusted by controlling the current intensity of the accelerator to adjust the intensity of the beam line.
[0040] The beam delivery system parameters may include vacuum parameters, position parameters of the reflector, monochromator and focusing mirror, and angle parameters of the reflector, monochromator and focusing mirror, etc. Through the adjustment of the beam delivery system, the beam can reach the target experimental area.
[0041] Among them, the experimental platform parameters are mainly the parameters of the detector. The detector parameters can include resolution and sensitivity, as well as the detector position. Specifically, the detector position can be adjusted by adjusting the position and angle of the detector to optimize the detection signal. The resolution and sensitivity can be achieved by selecting appropriate detectors and data acquisition systems to meet the requirements of time resolution and sensitivity of the experiment.
[0042] Furthermore, the state of the current beamline station, that is, the current device state, can be determined based on the device description parameters of the beamline station (specifically, beamline source parameters, beamline delivery system parameters, and laboratory table parameters). Specifically, the state description of the beamline station is generally generated by the radius and position of the spot (assuming that the desired spot is circular). It can be understood that different device description parameters of the beamline station can correspond to different states of the beamline station, that is, generate different spot radii and positions.
[0043] For example, assume that the current equipment status of the beamline station is , where the spot radius is and the location is . Current device status The corresponding device description parameters to be adjusted are ,in, represents the beam source energy; Indicates the beam source current intensity; Indicates vacuum degree parameter; Represents the position parameters of the reflector, monochromator and focusing mirror; Indicates the angle parameters of the reflector, monochromator and focusing mirror; Indicates resolution and sensitivity; Indicates the detector position.
[0044] S120: Obtain an expected device state that the target device expects to obtain.
[0045] The expected device state may be a device state that the current target device wants to achieve.
[0046] Continuing from the previous example, assume that the desired equipment state of the beamline station is , where the spot radius is and the location is .
[0047] S130: jointly input the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the device description parameters to be adjusted, thereby obtaining adjusted device description parameters.
[0048] The device parameter adjustment policy model may be a policy model for adjusting device description parameters.
[0049] In this embodiment, the current device state and the expected device state need to be jointly input into the device parameter adjustment strategy model to perform parameter adjustment processing, and the adjusted device description parameters can be obtained through the device parameter adjustment strategy model.
[0050] It is understandable that the adjusted device description parameters are the device description parameters corresponding to the expected device state. That is, after the current device description parameters to be adjusted are updated to the adjusted device description parameters, the target device is also converted from the current device state to the expected device state. For some industrial equipment, the current device description parameters to be adjusted are dynamically adjusted by knowing the current device state and the expected device state, thereby improving the efficiency and accuracy of device parameter adjustment, and reducing the time cost and labor cost of device parameter adjustment.
[0051] Continuing from the previous example, since the current device status can be obtained as and the expected device state is , you can adjust the device description parameters to be adjusted in the device parameter adjustment strategy model After adjustment, the adjusted device description parameters can be obtained. It can be assumed that the adjusted device description parameters are .
[0052] It is understandable that when the device description parameters are adjusted to When , that is, the beamline station can generate a spot radius of and the location is In this way, the parameters can be adjusted through the device parameter adjustment strategy model according to the current device state and the expected device state, so that the target light spot can be generated according to the adjusted parameters, which can improve the efficiency and accuracy of device parameter adjustment.
[0053] S140: Feedback the adjusted device description parameters to complete the adjustment process of the device description parameters to be adjusted.
[0054] In this embodiment, after the adjusted device description parameters are determined to be obtained, the adjusted device description parameters need to be fed back to the user to ensure that the adjustment process operation of the device description parameters to be adjusted is completed.
[0055] Optionally, after jointly inputting the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the device description parameters to be adjusted and obtaining the adjusted device description parameters, it also includes: receiving parameter fine-tuning calibration instructions in real time; parsing the parameter fine-tuning calibration instructions, and fine-tuning and calibrating the adjusted device description parameters through a pre-built parameter fine-tuning calibration model to obtain device fine-tuning calibration description parameters; and outputting the device fine-tuning calibration description parameters as feedback to the user.
[0056] The parameter fine-tuning calibration instruction may be an instruction for fine-tuning and calibrating the adjusted device description parameters. The parameter fine-tuning calibration model may be a model for fine-tuning and calibrating the adjusted device description parameters. The device fine-tuning calibration description parameters may be the fine-tuning calibration description parameters obtained after the adjusted device description parameters are fine-tuned and calibrated.
[0057] In this embodiment, since the peripheral influencing factors of the device will change during operation, which will cause the inherent characteristics of the device to change, the device needs to be fine-tuned and calibrated when necessary. Therefore, it is necessary to receive parameter fine-tuning and calibration instructions in real time, and implement fine-tuning and calibration processing of the adjusted device description parameters through the parameter fine-tuning and calibration model, thereby further improving the accuracy of device parameter adjustment, and being able to perform parameter fine-tuning and calibration in a timely manner according to changes in peripheral influencing factors of the device.
[0058] Continuing from the previous example, if a parameter fine-tuning calibration instruction is received, the adjusted device description parameters are adjusted through the parameter fine-tuning calibration model. To perform fine-tuning calibration processing, the device fine-tuning calibration description parameters can be further obtained.
[0059] The technical solution of the embodiment of the present invention obtains the current device state of the target device and the device description parameters to be adjusted corresponding to the current device state; obtains the expected device state that the target device expects to obtain; inputs the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the current device description parameters to be adjusted to obtain the adjusted device description parameters; and feeds back the adjusted device description parameters to complete the adjustment of the current device description parameters to be adjusted. The present invention solves the problem of low efficiency and low accuracy caused by the large amount of manpower, material resources and time cost required for device parameter adjustment, improves the efficiency and accuracy of device parameter adjustment, and reduces time cost and manpower cost.
[0060] Embodiment 2
[0061] Figure 2a A detailed flowchart of a device parameter adjustment method is provided for the second embodiment of the present invention. This embodiment is optimized based on the above embodiments. In this embodiment, before obtaining the current device status of the target device and the device description parameters to be adjusted corresponding to the current device status, it also includes training of the device parameter adjustment strategy model.
[0062] S210. Acquire multiple historical device states, historical device description parameters corresponding to each of the historical device states, historical expected device states corresponding to each of the historical device states, and historical expected device description parameters corresponding to each of the historical expected device states through a pre-constructed real device simulation model.
[0063] Among them, the real device simulation model is a model obtained by simulating various types of real devices.
[0064] The historical device status may be the status of different types of real devices collected at historical moments, and the historical device description parameters may be the parameters describing the devices collected at historical moments.
[0065] The historical expected device state may be a device state that the device wants to achieve under the historical device state at a historical moment. The historical expected device description parameter may be an expected device description parameter corresponding to the historical expected device state.
[0066] It can be understood that by simulating various types of real equipment to obtain a real equipment simulation model, the collection of data parameters on the real equipment is avoided. The data parameters can be collected directly on the real equipment simulation model. On the premise of ensuring the accuracy of the collected data parameters, the safe use of the real equipment is guaranteed and the damage to the real equipment caused by the collection of data parameters is reduced.
[0067] Figure 2b A schematic diagram of the structure of a beamline station in a device parameter adjustment method is provided for the second embodiment of the present invention. Figure 2b The beamline station includes a beamline source, a beamline delivery system and an experimental table.
[0068] Specifically, a beam source can produce high-intensity and high-brightness beams, such as synchrotron radiation sources or accelerators. A beam delivery system can be a beamline end station that guides the radiation beam from the radiation source to the laboratory bench, including vacuum equipment, various optical elements, such as mirrors, lenses, or monochromators. The laboratory bench can be the place where the actual experiment is carried out, equipped with various detectors and analytical instruments to detect and analyze the behavior and characteristics of the sample under the radiation beam.
[0069] In addition, beamline stations are experimental devices used to study and apply various forms of radiation beams (such as X-rays, synchrotron radiation, ion beams or electron beams, etc.). They are usually used for basic research and applied research in the fields of physics, chemistry, biology or materials science. A beamline station is a complex system. From the radiation source to the final experimental table, a large number of equipment description parameters need to be adjusted to obtain the required equipment status.
[0070] For example, the beamline station needs to be simulated first to obtain a real device simulation model corresponding to the real device of the beamline station, so that data parameters can be collected through the real device simulation model, reducing the possibility of equipment damage to the beamline station caused by data parameter collection.
[0071] Furthermore, after constructing the real equipment simulation model corresponding to the beamline station, the real equipment simulation model can be used to obtain multiple historical equipment states, historical equipment description parameters corresponding to each of the historical equipment states, historical expected equipment states corresponding to each of the historical equipment states, and historical expected equipment description parameters corresponding to each of the historical expected equipment states.
[0072] It is understandable that the historical device state can generate the historical spot radius and historical position. The historical device description parameters can include the historical beam source energy, the historical beam source current intensity, the historical vacuum parameters, the historical reflector, the monochromator and the focusing mirror position parameters, the historical reflector, the monochromator and the focusing mirror angle parameters, the historical resolution and sensitivity and the historical detector position. Similarly, the specific parameters of the historical expected device state and the historical expected device description parameters are not repeated here.
[0073] S220, traverse each group of historical device states, historical device description parameters, historical expected device states and historical expected device description parameters in turn, and train the pre-built initial device parameter adjustment strategy model through a reinforcement learning strategy optimization method to obtain a trained device parameter adjustment strategy model.
[0074] The initial device parameter adjustment strategy model may be a strategy model for adjusting device parameters. The reinforcement learning strategy optimization method may be constructed based on a post-experience replay algorithm.
[0075] Specifically, the post-experience replay algorithm may include a reward function and a spatial action, and state-action reward trajectory data generated by the reward function and the spatial action.
[0076] Optionally, the method sequentially traverses each group of historical device states, historical device description parameters, historical expected device states, and historical expected device description parameters, and performs model training on a pre-constructed initial device parameter adjustment strategy model through a reinforcement learning strategy optimization method to obtain a trained device parameter adjustment strategy model, including: sequentially traversing each group of historical device states and historical expected device states, and obtaining each historical reward function and each historical spatial action through a reinforcement learning strategy optimization method; generating each historical adjustment strategy device parameter according to each group of historical reward functions, historical spatial actions, historical device states, historical device description parameters, historical expected device states, and historical expected device description parameters; and training a pre-constructed initial device parameter adjustment strategy model according to each historical adjustment strategy device parameter to obtain a trained device parameter adjustment strategy model.
[0077] The historical reward function can be a reward function for judging the quality of the device parameter adjustment model based on the historical device state and the historical expected device state. The historical spatial action can be a strategy function based on the historical device state and the historical expected device state, and based on the use of any state. To calculate the spatial action.
[0078] In this embodiment, the historical adjustment strategy device parameters can be generated through the historical reward function, historical spatial action, historical device state, historical device description parameters, historical expected device state and historical expected device description parameters, and the initial device parameter adjustment strategy model can be trained based on the historical adjustment strategy device parameters. Specifically, through the reinforcement learning strategy optimization method, according to the formula , to get each historical reward function ;in, For historical equipment status; is the historical expected equipment state; MAE is the mean absolute error.
[0079] In addition, according to the formula , to get each historical spatial action a; among them, It is the reinforcement learning policy function in the hindsight experience replay algorithm.
[0080] In this embodiment, after determining the historical reward function and the historical spatial action, it is necessary to generate the historical adjustment strategy device parameters. Specifically, first, according to the strategy function By collecting states, spatial actions and reward functions, the state-action-reward trajectory data is generated. Assume that the initial historical device state is , then the state action reward trajectory data is ,in, is the historical device state corresponding to the next historical space action, represents the next historical space action, Represents the historical reward function corresponding to the next historical space action. is the historical device state corresponding to the next m historical space actions, represents the next m historical spatial actions, Represents the historical reward function corresponding to the next m historical space actions.
[0081] Furthermore, the parameters of each historical adjustment strategy device can be obtained based on the state action reward trajectory data. The parameters of the historical adjustment strategy device parameters are obtained by using the following format: represents the i-th historical device state. The state-action reward trajectory data may include m+1 historical device states, that is, the value range of i is greater than or equal to 1 and less than or equal to m+1. For example, when i=1, it can be determined ; When i=2, it can be determined . express The historical device state corresponding to the next historical space action.
[0082] Correspondingly, the device parameters of each group of historical adjustment strategies can be obtained , it can be put into the experience replay pool, and the device parameter adjustment model can be retrained through each historical adjustment strategy device parameter to obtain a trained device parameter adjustment strategy model.
[0083] S230: Acquire the current device state of the target device and the device description parameters to be adjusted corresponding to the current device state.
[0084] S240: Obtain an expected device state that the target device expects to obtain.
[0085] S250: jointly input the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the device description parameters to be adjusted, thereby obtaining adjusted device description parameters.
[0086] S260: Feedback the adjusted device description parameters to complete the adjustment process of the device description parameters to be adjusted.
[0087] Optionally, after the pre-constructed initial device parameter adjustment strategy model is trained according to each historical adjustment strategy device parameter to obtain a trained device parameter adjustment strategy model, it also includes: if it is determined that the operating environment corresponding to various types of real devices has changed, then on each real device, each historical device fine-tuning state, and the historical device fine-tuning description parameters corresponding to each of the historical device fine-tuning states, the historical expected device fine-tuning states corresponding to each of the historical device fine-tuning states, and the historical expected device fine-tuning description parameters corresponding to each of the historical expected device fine-tuning states are collected; each group of historical device fine-tuning states, historical device fine-tuning description parameters, historical expected device fine-tuning states and historical expected device fine-tuning description parameters are traversed in turn, and the device parameter adjustment strategy model is retrained through the reinforcement learning strategy optimization method to obtain a trained parameter fine-tuning calibration model.
[0088] In this embodiment, since the peripheral influencing factors in the operating environment will change during the operation of the device, which will cause the inherent characteristics of the device to change, the device needs to be fine-tuned and calibrated when necessary. Therefore, it is necessary to obtain the historical device fine-tuning status, historical device fine-tuning description parameters, historical expected device fine-tuning status and historical expected device fine-tuning description parameters on each real device, and train the parameter fine-tuning calibration model through the reinforcement learning strategy optimization method.
[0089] Continuing with the previous example, in this embodiment, if it is determined that the operating environment corresponding to various types of real equipment has changed, assuming that the operating environment of the beam line station has changed, it is necessary to collect parameters on the beam line station.
[0090] Specifically, the fine-tuning status of each historical device, the fine-tuning description parameters of each historical device, the fine-tuning status of each historical expected device, and the fine-tuning description parameters of each historical expected device can be collected at the beamline station to retrain the device parameter adjustment strategy model and obtain a trained parameter fine-tuning calibration model. In this way, the fine-tuning and calibration of parameters can be carried out in a timely manner according to the changes in the actual device operating environment, which can improve the accuracy of device parameter adjustment.
[0091] Furthermore, after the parameter fine-tuning and calibration model is trained, when a parameter fine-tuning and calibration instruction is received, the parameter fine-tuning and calibration instruction can be parsed, and the adjusted device description parameters can be fine-tuned and calibrated through the parameter fine-tuning and calibration model to obtain the device fine-tuning and calibration description parameters, and feedback is provided, thereby improving the flexibility of device parameter adjustment and better improving the accuracy of device parameter adjustment.
[0092] The technical solution of the embodiment of the present invention uses a pre-constructed real device simulation model to obtain multiple historical device states, historical device description parameters corresponding to each of the historical device states, historical expected device states corresponding to each of the historical device states, and historical expected device description parameters corresponding to each of the historical expected device states; traverses each group of historical device states, historical device description parameters, historical expected device states, and historical expected device description parameters in turn, and uses a reinforcement learning strategy optimization method to train the pre-constructed initial device parameter adjustment strategy model to obtain a trained device parameter adjustment strategy model; obtains the current device state of the target device, and the device description parameters to be adjusted corresponding to the current device state; obtains the expected device state that the target device expects to obtain; jointly inputs the current device state and the expected device state into the pre-trained device parameter adjustment strategy model to adjust the current device description parameters to be adjusted to obtain the adjusted device description parameters; and feeds back the adjusted device description parameters to complete the adjustment of the current device description parameters to be adjusted. A more accurate device parameter adjustment strategy model is trained, and the device parameter adjustment strategy model can improve the efficiency and accuracy of device parameter adjustment, and reduce time and labor costs.
[0093] Embodiment 3
[0094] Figure 3 This is a schematic diagram of the structure of a device parameter adjustment device provided in Embodiment 3 of the present invention. The device parameter adjustment device provided in this embodiment can be implemented by software and / or hardware, and can be configured in a terminal device or a server to implement a device parameter adjustment method in an embodiment of the present invention. Figure 3 As shown, the apparatus includes: a current device state and device description parameter acquisition module 310, an expected device state acquisition module 320, an adjusted device description parameter determination module 330 and an adjusted device description parameter feedback module 340.
[0095] The current device state and device description parameter acquisition module 310 is used to acquire the current device state of the target device and the device description parameters to be adjusted corresponding to the current device state; wherein the target device is an artificial intelligence-driven device;
[0096] The expected device state acquisition module 320 is used to acquire the expected device state that the target device expects to obtain;
[0097] The adjusted device description parameter determination module 330 is used to input the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the current device description parameters to be adjusted, so as to obtain the adjusted device description parameters;
[0098] The adjusted device description parameter feedback module 340 is used to feed back the adjusted device description parameters to complete the adjustment process of the device description parameters to be adjusted.
[0099] The technical solution of the embodiment of the present invention obtains the current device state of the target device and the device description parameters to be adjusted corresponding to the current device state; obtains the expected device state that the target device expects to obtain; inputs the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the current device description parameters to be adjusted to obtain the adjusted device description parameters; and feeds back the adjusted device description parameters to complete the adjustment of the current device description parameters to be adjusted. The present invention solves the problem of low efficiency and low accuracy caused by the large amount of manpower, material resources and time cost required for device parameter adjustment, improves the efficiency and accuracy of device parameter adjustment, and reduces time cost and manpower cost.
[0100] On the basis of the above embodiments, it also includes a parameter fine-tuning calibration instruction receiving module, which can be specifically used for: after the current device state and the expected device state are jointly input into a pre-trained device parameter adjustment strategy model to adjust the device description parameters to be adjusted and the adjusted device description parameters are obtained, parameter fine-tuning calibration instructions are received in real time; the parameter fine-tuning calibration instructions are parsed, and the adjusted device description parameters are fine-tuned and calibrated by a pre-built parameter fine-tuning calibration model to obtain device fine-tuning calibration description parameters; and the device fine-tuning calibration description parameters are fed back to the user.
[0101] On the basis of the above embodiments, it also includes a device parameter adjustment strategy model training module, which can be specifically used for: before obtaining the current device state of the target device and the device description parameters to be adjusted corresponding to the current device state, a plurality of historical device states, historical device description parameters corresponding to each of the historical device states, historical expected device states corresponding to each of the historical device states, and historical expected device description parameters corresponding to each of the historical expected device states are obtained through a pre-constructed real device simulation model; wherein the real device simulation model is a model obtained by simulating various types of real devices; each group of historical device states, historical device description parameters, historical expected device states and historical expected device description parameters are traversed in turn, and the pre-constructed initial device parameter adjustment strategy model is trained through a reinforcement learning strategy optimization method to obtain a trained device parameter adjustment strategy model.
[0102] On the basis of the above embodiments, the device parameter adjustment strategy model training module can also be specifically used for: traversing each group of historical device states and historical expected device states in turn, and obtaining each historical reward function and each historical space action through a reinforcement learning strategy optimization method; generating each historical adjustment strategy device parameter according to each group of historical reward functions, historical space actions, historical device states, historical device description parameters, historical expected device states, and historical expected device description parameters; training a pre-constructed initial device parameter adjustment strategy model according to each historical adjustment strategy device parameter to obtain a trained device parameter adjustment strategy model.
[0103] On the basis of the above embodiments, the reinforcement learning strategy optimization method is constructed based on the post-experience replay algorithm.
[0104] On the basis of the above embodiments, the device parameter adjustment strategy model training module can also be specifically used to: through the reinforcement learning strategy optimization method, according to the formula , to get each historical reward function ;in, For historical equipment status; is the historical expected device state; MAE is the mean absolute error; according to the formula , to get each historical spatial action a; among them, It is the reinforcement learning policy function in the hindsight experience replay algorithm.
[0105] On the basis of the above embodiments, it also includes a parameter fine-tuning calibration model determination module, which can be specifically used for: after the pre-constructed initial device parameter adjustment strategy model is trained according to the device parameters of each historical adjustment strategy to obtain a trained device parameter adjustment strategy model, if it is determined that the operating environment corresponding to various types of real devices has changed, then on each real device, each historical device fine-tuning state, and the historical device fine-tuning description parameters corresponding to each of the historical device fine-tuning states, the historical expected device fine-tuning states corresponding to each of the historical device fine-tuning states, and the historical expected device fine-tuning description parameters corresponding to each of the historical expected device fine-tuning states are collected; each group of historical device fine-tuning states, historical device fine-tuning description parameters, historical expected device fine-tuning states and historical expected device fine-tuning description parameters are traversed in turn, and the device parameter adjustment strategy model is retrained through the reinforcement learning strategy optimization method to obtain a trained parameter fine-tuning calibration model.
[0106] The device parameter adjustment apparatus provided in the embodiment of the present invention can execute the device parameter adjustment method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.
[0107] Embodiment 4
[0108] Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement the fourth embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.
[0109] like Figure 4As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. The input / output (I / O) interface 15 is also connected to the bus 14.
[0110] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0111] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a device parameter adjustment method.
[0112] In some embodiments, the device parameter adjustment method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the device parameter adjustment method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the device parameter adjustment method in any other appropriate manner (e.g., by means of firmware).
[0113] The method includes: obtaining the current device state of the target device and the device description parameters to be adjusted corresponding to the current device state; wherein the target device is a device driven by artificial intelligence; obtaining the expected device state that the target device expects to obtain; inputting the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the current device description parameters to be adjusted to obtain the adjusted device description parameters; and feeding back the adjusted device description parameters to complete the adjustment of the current device description parameters to be adjusted.
[0114] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0115] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0116] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0117] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).
[0118] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.
[0119] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.
[0120] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.
[0121] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
[0122] Embodiment 5
[0123] Embodiment 5 of the present invention also provides a computer-readable storage medium, wherein the computer-readable instructions are used to execute a device parameter adjustment method when executed by a computer processor, the method comprising: obtaining a current device state of a target device, and a device description parameter to be adjusted corresponding to the current device state; wherein the target device is an artificial intelligence-driven device; obtaining an expected device state that the target device expects to obtain; jointly inputting the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the current device description parameter to be adjusted to obtain the adjusted device description parameter; and feeding back the adjusted device description parameter to complete the adjustment of the current device description parameter to be adjusted.
[0124] Of course, the computer-readable storage medium provided in the embodiment of the present invention has computer-executable instructions that are not limited to the method operations described above, and can also execute related operations in the device parameter adjustment method provided in any embodiment of the present invention.
[0125] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.
[0126] It is worth noting that in the embodiment of the above-mentioned device parameter adjustment apparatus, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.
[0127] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for adjusting device parameters, characterized in that: include: Acquire a current device state of a target device and a device description parameter to be adjusted corresponding to the current device state; wherein the target device is an artificial intelligence-driven device; Get the expected device state that the target device expects to obtain; The current device state and the expected device state are jointly input into a pre-trained device parameter adjustment strategy model to adjust the device description parameters to be adjusted, thereby obtaining the adjusted device description parameters; Feedback the adjusted device description parameters to complete the adjustment process of the device description parameters to be adjusted; Before acquiring the current device state of the target device and the device description parameter to be adjusted corresponding to the current device state, the method further includes: Acquire multiple historical device states, historical device description parameters corresponding to each of the historical device states, historical expected device states corresponding to each of the historical device states, and historical expected device description parameters corresponding to each of the historical expected device states through a pre-built real device simulation model; Among them, the real device simulation model is a model obtained by simulating various types of real devices; Each group of historical device states, historical device description parameters, historical expected device states, and historical expected device description parameters are traversed in turn, and the pre-built initial device parameter adjustment strategy model is trained by using a reinforcement learning strategy optimization method to obtain a trained device parameter adjustment strategy model; The device description parameters to be adjusted are device description parameters collected from different dimensions; each current device state corresponds to a set of device description parameters.
2. The method according to claim 1, characterized in that After the current device state and the expected device state are jointly input into a pre-trained device parameter adjustment strategy model to adjust the device description parameters to be adjusted, and the adjusted device description parameters are obtained, the method further includes: Receive parameter fine-tuning and calibration instructions in real time; Parsing the parameter fine-tuning calibration instruction, and performing fine-tuning calibration processing on the adjusted device description parameters through a pre-built parameter fine-tuning calibration model to obtain device fine-tuning calibration description parameters; The device is fine-tuned and calibrated to describe parameters and output feedback to the user.
3. The method according to claim 2, characterized in that The method sequentially traverses each group of historical device states, historical device description parameters, historical expected device states, and historical expected device description parameters, and trains the pre-built initial device parameter adjustment strategy model through a reinforcement learning strategy optimization method to obtain a trained device parameter adjustment strategy model, including: Traverse each set of historical device states and historical expected device states in turn, and obtain each historical reward function and each historical spatial action through the reinforcement learning strategy optimization method; Generate each historical adjustment strategy device parameter according to each set of historical reward functions, historical spatial actions, historical device states, historical device description parameters, historical expected device states, and historical expected device description parameters; The pre-built initial device parameter adjustment strategy model is trained according to each historical adjustment strategy device parameter to obtain a trained device parameter adjustment strategy model.
4. The method according to claim 3, characterized in that The reinforcement learning strategy optimization method is constructed based on a post-experience replay algorithm.
5. The method according to claim 4, characterized in that The method of sequentially traversing each group of historical device states and historical expected device states and obtaining each historical reward function and each historical spatial action by using a reinforcement learning strategy optimization method includes: Through the reinforcement learning strategy optimization method, according to the formula , to get each historical reward function ; in, For historical device status; is the historical expected equipment state; MAE is the mean absolute error; According to the formula , to get each historical spatial action a; in, It is the reinforcement learning policy function in the hindsight experience replay algorithm.
6. The method according to claim 5, characterized in that After the pre-built initial device parameter adjustment strategy model is trained according to each historical adjustment strategy device parameter to obtain a trained device parameter adjustment strategy model, the method further includes: If it is determined that the operating environment corresponding to various types of real devices has changed, then on each real device, each historical device fine-tuning state, and the historical device fine-tuning description parameters corresponding to each of the historical device fine-tuning states, the historical expected device fine-tuning state corresponding to each of the historical device fine-tuning states, and the historical expected device fine-tuning description parameters corresponding to each of the historical expected device fine-tuning states are collected; Each group of historical device fine-tuning states, historical device fine-tuning description parameters, historical expected device fine-tuning states, and historical expected device fine-tuning description parameters are traversed in turn, and the device parameter adjustment strategy model is retrained through the reinforcement learning strategy optimization method to obtain a trained parameter fine-tuning calibration model.
7. A device for adjusting equipment parameters, characterized in that: include: A current device state and device description parameter acquisition module, used to acquire the current device state of a target device and the device description parameters to be adjusted corresponding to the current device state; wherein the target device is an artificial intelligence-driven device; An expected device state acquisition module is used to acquire an expected device state that a target device expects to obtain; An adjusted device description parameter determination module is used to jointly input the current device state and the expected device state into a pre-trained device parameter adjustment strategy model to adjust the current device description parameters to be adjusted, thereby obtaining adjusted device description parameters; The adjusted device description parameter feedback module is used to feed back the adjusted device description parameter to complete the adjustment process of the device description parameter to be adjusted; It also includes a device parameter adjustment strategy model training module for: Before obtaining the current device state of the target device and the device description parameters to be adjusted corresponding to the current device state, a plurality of historical device states, historical device description parameters corresponding to each of the historical device states, historical expected device states corresponding to each of the historical device states, and historical expected device description parameters corresponding to each of the historical expected device states are obtained through a pre-constructed real device simulation model; wherein the real device simulation model is a model obtained by simulating various types of real devices; each group of historical device states, historical device description parameters, historical expected device states, and historical expected device description parameters are traversed in turn, and a pre-constructed initial device parameter adjustment strategy model is trained through a reinforcement learning strategy optimization method to obtain a trained device parameter adjustment strategy model; The device description parameters to be adjusted are device description parameters collected from different dimensions; each current device state corresponds to a set of device description parameters.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the device parameter adjustment method according to any one of claims 1 to 6 is implemented.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the device parameter adjustment method according to any one of claims 1 to 6 when executed.
