CN118824194A - A backlight source control system and method for LED liquid crystal screen - Google Patents

Abstract

The invention relates to the field of intelligent control, and discloses a backlight source control system and method of an LED liquid crystal screen, wherein the control system comprises the following components: the control system and the control method have important significance in solving the situations of power consumption management control, heat dissipation management control, improving the user experience of display equipment, saving energy, prolonging the service life of an LED backlight source and the like.

Description

A backlight source control system and method for LED liquid crystal screen

Technical Field

The invention patent relates to the field of intelligent control, and more specifically, to a backlight source control system and method for an LED liquid crystal screen.

Background Art

LED backlight is an indispensable component in display devices such as liquid crystal displays, used to provide background light to increase brightness and contrast. Traditional LED backlight control solutions are usually based on fixed preset brightness settings or simple manual adjustments, which cannot be flexibly adaptively adjusted according to actual outdoor scene requirements and lack effective power consumption management control.

In addition, long-term operation can easily cause a series of problems such as burnout of components due to overheating. Although the current LED backlight control system can control the on and off of the LED backlight by controlling the on and off of the circuit, the previous control system cannot adjust the heat dissipation speed of the LED backlight, which can easily damage the LED backlight due to untimely heat dissipation during the use of the LED backlight.

Traditionally, power management control and heat management control are dispersed into different modules. It is expected that an optimized LED backlight control system method can use a set of control structures to simultaneously meet the needs of user-defined backlight brightness management, intelligent adaptive power management, and avoiding high temperature overload. This is of great significance for improving the user experience of display devices, saving energy, and increasing the life of LED backlight sources. For this reason, we propose a backlight control system and method for LED LCD screens.

Summary of the invention

The object of the present invention is to provide a backlight source control system and method for an LED liquid crystal screen to solve the problems raised in the above background technology.

To achieve the above object, the present invention provides the following technical solution: a backlight source control system for an LED liquid crystal screen, comprising:

Multi-sensory information collection module, used to detect and collect a series of sensory information closely related to common outdoor scenes;

The control module can communicate and receive information from the information acquisition module, generate a feedback control signal and connect it to act on the drive circuit module;

A driving circuit module receives a feedback signal generated by the control module and drives the brightness and tone calling module;

A brightness and tone adjustment module, for determining, based on the generated feedback signal, whether the brightness value of the LED backlight source should be increased, maintained, or decreased, or whether the tone value should be kept bright or dim;

The backlight module is used to receive adjustment information and make corresponding adjustments.

Preferably, the multi-sensory information acquisition module includes:

Used to receive a user-defined user interface;

An ambient light sensor for receiving ambient light intensity;

Cameras for receiving pedestrian or vehicle traffic;

Temperature sensor for receiving device temperature.

Preferably, the control module comprises:

An intelligent computing decision unit for receiving sensory information and generating control signals based on intelligent algorithms;

A rule queue generation mechanism for generating multi-level rule queues;

A mechanism for transmitting control signals to drive circuits.

Preferably, the driving circuit module and the brightness and tone adjustment module can receive a control signal to drive the control and adjustment of the brightness and tone of the backlight source.

A backlight source control method for an LED liquid crystal screen comprises the following steps:

Step 1: After the perception information is collected and cleaned and feature abstracted, the data is divided according to the dimension of the feature vector and applied to different intelligent computing units. The high-dimensional data uses the neural network model for feature extraction and discrimination, and the low-dimensional data uses the fuzzy system for feature extraction and discrimination;

Step 2: To ensure that multi-view and multi-modal data can collaboratively learn complementary and efficient features, the gradient update adopts a mechanism similar to that of the parameter server;

Step 3: Specifically, it is divided into two parts, in which each local intelligent computing unit only calculates the data set related to itself and generates local gradients. These gradient information are aggregated to a common node and accumulated. Then, this common node updates the overall weight parameter according to the common gradient, and distributes this overall parameter to different local intelligent computing units according to the corresponding relationship;

Step 4: After reaching the desired error, the four local intelligent computing units generate four exclusive feedback generation signals, which are further used to generate corresponding rule queues using fuzzy synthesis.

Preferably, the multi-level rule queue generation mechanism includes: a three-level rule queue based on a set mode, which consists of three modes: user-defined mode, performance mode, and power-saving mode, wherein the user-defined mode is used for the user's detailed customization needs, the high-performance mode is used when the power supply is sufficient, and the power-saving mode is used for the unstable power supply mode using the solar energy + battery module; the synthesis method of the mode uses a hierarchical fuzzy system, and the hierarchical fuzzy system consists of a hierarchical synthesis method of "primary feedback synthesis", "secondary feedback synthesis" and "final feedback synthesis".

Preferably, the steps of the three modes of user-defined mode, performance mode and power saving mode in the multi-level rule queue generation mechanism are as follows:

Power saving mode:

The main feedback synthesis is first fuzzy synthesized by "user-defined information" and "temperature information". This design is because user-defined information and temperature information play the most important role in power saving mode;

Secondary feedback synthesis is a fuzzy synthesis of "ambient light information" and "pedestrian and vehicle flow information (which can be collected by sound sensors or cameras, etc.)". The reason why it is secondary feedback information is that their feedback results are only used to fine-tune the results of the primary feedback;

The final feedback synthesis is used to generate the final feedback signal. Its synthesis logic is a final synthesis of the main feedback generation signal and the negative feedback generation signal. The final synthesis result is a pile of scalar information.

Performance Mode:

The primary feedback synthesis is fuzzy synthesized by "user-defined information" and "ambient light information"; the secondary feedback synthesis is synthesized by "temperature information" and "pedestrian and vehicle flow information"; the final synthesis is further synthesized by the results of the primary feedback information and the secondary feedback information;

User defined mode:

The primary feedback consists of "user-defined information" as the only input; the secondary feedback consists of three inputs: "ambient light information", "pedestrian and vehicle flow information" and "temperature information"; and the final synthesis of the primary feedback information and the secondary feedback information is further synthesized.

Compared with the prior art, the control system and method provided by the present invention have the following beneficial effects:

1. The control system and method receive a variety of information from the information acquisition module, provide an integrated customizable, scalable, and adaptive control method for multiple conditions, and provide on-demand backlight power consumption management and intelligent adjustment capabilities.

2. This control system and method can not only ensure better visual effects and effectively reduce the energy consumption of the LED backlight source, but also effectively prevent the backlight source from being damaged due to untimely heat dissipation during use, thereby increasing the service life of the LED backlight source.

3. The control system and method provide a multi-perspective, multi-modal, and multi-intelligent representation multi-sensory information fusion mechanism, and generate a preloadable multi-level rule queue through the collaborative learning of these perceptual information. While shortening the control link, it separates training from reasoning, effectively saves resources, reduces power consumption, and improves the robustness and stability of the system.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the present application and, together with the description, serve to explain the principles of the present application.

FIG1 shows an overall block diagram of an LED backlight control system according to an embodiment of the present application;

FIG2 shows a framework of a control module shown in an LED backlight control system according to an embodiment of the present application;

FIG. 3 shows an example of fuzzy synthesis for generating a multi-level rule queue of a control module shown in an LED backlight control system according to an embodiment of the present application.

DETAILED DESCRIPTION

The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

Various exemplary embodiments, features and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

In addition, in order to better illustrate the present application, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present application can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present application.

Fig. 1 shows a schematic block diagram of the overall backlight source control system of the LED liquid crystal screen according to an embodiment of the present application. As shown in Fig. 1, it includes an information acquisition module, a control module, a drive circuit module, a brightness control module, and a backlight source group.

Specifically, the information collection module includes and processes the following scenarios: ① When the ambient light changes from bright to dark or from dark to bright, the advertising screen has a corresponding adaptive brightness adjustment function. The collection method is the ambient light sensor. ② When there is a lot of people and traffic at night, the advertising screen should have a striking and bright brightness. The collection method is a camera or sound sensor or field survey data. ③ When long-term work causes huge heat generation of the device, the advertising screen should have appropriate adaptive power consumption reduction to avoid long-term high temperature overload causing device burnout. The collection method is a temperature sensor. ④ When the user wants to do fine brightness control by himself, the advertising screen should be able to meet the user's customization needs. The collection method is the user interface.

Specifically, the control module is used to process data samples such as cameras/sensors/field data that have different perspectives but are inherently related in some sense, so that they can be effectively integrated and generate corresponding control signals. It adopts a multi-perspective collaborative learning mechanism to generate the final control rule queue.

Correspondingly, as shown in Figure 2, according to the needs of the control target, the multi-sensory information acquisition module is used to collect the corresponding information and summarize the corresponding perspective data set. In general, these data sets are divided into two categories after processing: high-dimensional data, such as a 28*28 picture, which can be regarded as a 784-dimensional data. The implementation method is mainly a neural network that integrates dimensionality reduction and feature extraction. Low-dimensional data, mainly sensor or statistical sample data, are characterized by a single or limited dimension, lack of accuracy and sufficient satisfaction. This kind of data is more suitable for using fuzzy system as a feature extraction and decision-making method. is a feature vector abstracted from a low-dimensional or high-dimensional data set, f is a neural network model or a fuzzy system model, and y is the output result of the data sample in the corresponding model.

Specifically, the fusion of the neural network model or the fuzzy system model is reflected in the use of a collaborative learning mechanism of multiple perspectives. Different from the traditional definition of "single sample from different perspectives", the concept of multiple perspectives is expanded in this implementation, which refers to different decision-making methods under the same control goal. Collaborative learning means that the optimization mechanism of the multi-perspective model adopts joint training rather than combining them after each training. For the four perspectives described in this embodiment, the formula of the joint training mechanism is shown in (1):

L＝L ₁ +L ₂ +L ₃ +L ₄ (1)

Among them, L is the overall optimization loss function, and _L1 ~ _L4 are the loss functions in the corresponding intelligent modes based on perspectives 1 to 4. The joint optimization mechanism allows the intelligent computing decision unit to make full use of the diversity and complementarity of data sample representations under different perspectives to construct the global optimal customized decision unit parameters.

Specifically, to further refine the formula (1), first, the optimization mechanism of a single intelligent control algorithm generally has the optimization mechanism described in formula (2). Based on the perception information collected by the information collection module, it is abstracted into data individuals. is the input of the sample individual, usually in the form of a high-dimensional vector. _yi is the output of the sample individual, usually in the form of a scalar. w is the trained parameter of the neural network model or fuzzy system model. Ω is the regularization term that controls the fuzzy complexity.

when When the data collection sources are heterogeneous, the above can be split into optimization objectives under four perspectives as shown in this implementation. The heterogeneous data set part can be processed by different control methods in the intelligent computing decision unit. In this implementation, the main processing is based on dimension. The information from the picture is processed by the neural network module, and the information from the sensor part is processed by the fuzzy system.

Combined with formula (1), formula (2) is further divided into the form described in formula (3) according to the heterogeneous data set, where l ₁ ~l ₄ are the optimized forms after splitting.

When the data set is split into different intelligent algorithms, each control algorithm actually only gets a part of the data set and weight information. In order to take advantage of collaborative learning, the training of formula (3) adopts the parameter server mechanism, that is, each control algorithm only uses the corresponding perception information, and then calculates the gradient information related to the perception information it has. However, the update of weights needs to be aggregated to a common parameter server for global weight update. The gradient update calculation process of each individual l is shown in formula (4). Among them, g _r ^(t) is the gradient information calculated based on its corresponding data set and weight information at time t. _{w r} ^(t) is the corresponding local weight information at time t.

The parameter server summarizes the gradients of the four objective optimization functions, calculates the total gradient, and then uses the gradient descent algorithm to calculate the new round of gradients. Each control algorithm then takes out the corresponding one and continues to complete the above update. The calculation process is shown in formula (5). Where g ^(t) is the global gradient at time t, and w ^(t) is the global weight information at time t.

The control algorithm for each viewing angle generates a corresponding feedback signal. In this embodiment, a total of four control signals are generated.

Accordingly, the feedback signals generated by the intelligent computing and decision-making unit can be further integrated using a fuzzy system to achieve further accurate control of the backlight source such as color tone and power consumption.

In this implementation, the rule queue is composed of three levels of teams, and the decision is selected by polling the rule queue and setting the corresponding mode based on the priority. Among them, the rule 1 queue has the highest priority, which corresponds to the setting with the highest weight for the user mode, while allowing other parameters to be modified to a certain extent. Different correction levels constitute different individuals in the rule 1 queue. The rule 2 queue has a suboptimal priority, which corresponds to the optimal system selected by the user under the default setting conditions by the decision optimization formula (1), which can be considered as a "high performance mode/gorgeous mode" for use under stable power supply conditions. Similarly, different correction levels also constitute different individuals in the rule 2 queue. Compared with rule 2, rule 3 mainly focuses on the "power saving mode" and can be used in outdoor conditions such as roads + when the power supply is unstable when using solar energy or batteries.

Taking "power saving mode" as an example, the fuzzy rule synthesis strategy for the rule queue generation mechanism is as follows: ① Main feedback synthesis: "user-defined information" and "temperature information" are first fuzzy synthesized. This design is because user-defined information and temperature information play the most important role in power saving mode. ② Secondary feedback synthesis: "ambient light information" and "people flow and vehicle flow information (can be sound sensor or camera acquisition, etc.)" are fuzzy synthesized. The reason why it is secondary feedback information is that their feedback results are only used to fine-tune the results of the main feedback. ③ Final feedback synthesis: used to generate the final feedback signal, its synthesis logic is the final synthesis of the main feedback generation signal and the negative feedback generation signal, and the final synthesis result is a pile of scalar information. The overall synthesis logic is shown in Figure 3. Each level of the rule queue only uses the same fuzzy synthesis logic, but allows the generation of multiple control values according to the degree of control looseness.

The driving circuit module and the brightness control module described in this embodiment are driven by the control signal generated by the control module and act on the brightness control module. With the help of feedback parameters, on the one hand, the transmittance can be adjusted to control the cold and warm color modes, and on the other hand, the power consumption can be adjusted to control the brightness, thereby controlling the different modes of the backlight source as a whole.

In the backlight module of this embodiment, the LED backlight source is composed of a plurality of LEDs to form an LED group. The LED backlight source is used to emit light of different brightness and hue.

Claims (7)

1. A backlight control system for an LED liquid crystal display, comprising:

The multi-perception information acquisition module is used for detecting and acquiring a series of perception information closely related to an outdoor common scene;

the control module can be used for receiving the information from the information acquisition module in a communication way, generating a feedback control signal and connecting the feedback control signal with the driving circuit module;

the driving circuit module receives the feedback signal generated by the control module and drives the brightness tone calling module;

A brightness and shade adjustment module for determining, based on the generated feedback signal, that the brightness value of the LED backlight should be increased, should be maintained or should be decreased, or that the shade value should be maintained bright or dim;

and the backlight source module is used for receiving the adjustment information and performing corresponding adjustment.

2. The backlight control system of an LED liquid crystal display of claim 1, wherein said multi-sensor information acquisition module comprises:

A user interface for receiving user-definitions;

An ambient light sensor for receiving an ambient light intensity;

The camera is used for receiving the flow of people or the flow of vehicles;

A temperature sensor for receiving a temperature of the device.

3. The LED lcd backlight control system of claim 1, wherein the control module comprises:

the intelligent calculation decision unit is used for receiving the perception information and generating a control signal according to an intelligent algorithm;

a rule queue generating mechanism for generating a multi-level rule queue;

a mechanism for transmitting control signals to the drive circuit.

4. The system of claim 1, wherein the driving circuit module and the brightness and color tone adjusting module are capable of receiving control signals to control and adjust brightness and color tone of the backlight.

5. The backlight control method of the LED liquid crystal screen is characterized by comprising the following steps of:

Step 1: after cleaning and feature abstraction, the sensing information acquisition data are respectively applied to different intelligent computing units according to dimension division of feature vectors, wherein high-dimensional data are subjected to feature extraction and discrimination by using a neural network model, and low-dimensional data are subjected to feature extraction and discrimination by using a fuzzy system;

step 2: in order to ensure that the multi-view multi-mode data can cooperatively learn complementary efficient characteristics, the gradient is updated by adopting a mechanism similar to a parameter server;

step 3: in particular, it is divided into two parts, in which each local intelligent computing unit only calculates the data set related to itself and generates local gradients, and these gradient information are summarized to a common node and accumulated, then the common node updates the overall weight parameters according to the common gradient, and distributes the overall parameters to different local intelligent computing units according to the corresponding relation;

Step 4: after the expected error is reached, the four local intelligent computing units generate 4 exclusive feedback generation signals, and the 4 feedback generation signals further generate a corresponding rule queue by using fuzzy synthesis.

6. The method for controlling a backlight of an LED liquid crystal display according to claim 5, wherein the multi-stage rule queue generating mechanism comprises: the three-level rule queue based on the set mode consists of three modes, namely a user-defined mode, a performance mode and a power-saving mode, wherein the user-defined mode is used for fine customization requirements of users, the high-performance mode is used for adopting an unstable power supply mode of a solar energy and storage battery module under the condition of sufficient power supply, the mode synthesis mode uses a level fuzzy system, and the level fuzzy system consists of a level synthesis mode of 'main feedback synthesis', 'sub feedback synthesis', 'final feedback synthesis'.

7. The method for controlling backlight source of LED liquid crystal display according to claim 6, wherein the steps of the user-defined mode, the performance mode and the power saving mode in the multi-stage rule queue generating mechanism are as follows,

Power saving mode:

the main feedback synthesis is that the user-defined information and the temperature information are firstly subjected to fuzzy synthesis, and the design is that the user-defined information and the temperature information play the most important weight in the power saving mode;

Sub-feedback synthesis, in which 'ambient light information' and 'people flow and traffic flow information (which can be acquired by a sound sensor or a camera, etc)' are subjected to fuzzy synthesis, wherein the sub-feedback information is because the feedback results of the sub-feedback information are only the results for fine tuning main feedback;

Final feedback synthesis, for generating a final feedback signal, wherein the synthesis logic performs final synthesis by the main feedback generation signal and the negative feedback generation signal, and the final synthesis result is a pile of scalar information;

Performance mode:

The system comprises a main feedback synthesis part, a secondary feedback synthesis part, a final synthesis part and a feedback control part, wherein the main feedback synthesis part performs fuzzy synthesis by user-defined information and environment light information, and the secondary feedback synthesis part performs synthesis by temperature information and traffic flow information;

User-defined modes:

The main feedback is composed of 'user-defined information' as unique input, the secondary feedback is composed of three inputs of 'ambient light information', 'traffic flow, traffic flow information' and 'temperature information', and the result of finally synthesizing the main feedback information and the secondary feedback information is further synthesized.